1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.450 2007/01/02 11:25:00 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
10 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
13 MH/01 Don't check that the operands of numeric comparison operators are
14 integers when their expansion is in "skipping" mode (fixes bug
15 introduced by 4.64-PH/07).
17 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
18 child addresses, Exim now panics and dies. Previously, because the count
19 is held in a short int, deliveries were likely to be lost. As such a
20 large number of recipients for a single message is ridiculous
21 (performance will be very, very poor), I have chosen to impose a limit
22 rather than extend the field.
28 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
29 leftover -K file (the existence of which was triggered by #402).
30 While we were at it, introduced process PID as part of the -K
31 filename. This should rule out race conditions when creating
34 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
35 processing considerably. Previous code took too long for large mails,
36 triggering a timeout which in turn triggers #401.
38 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
39 in the DK code in transports.c. sendfile() is not really portable,
40 hence the _LINUX specificness.
42 TF/01 In the add_headers option to the mail command in an Exim filter,
43 there was a bug that Exim would claim a syntax error in any
44 header after the first one which had an odd number of characters
47 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
48 callout verification, Exim cached a "reject" for the entire domain. This
49 is correct for most verifications, but it is not correct for a recipient
50 verification with use_sender or use_postmaster set, because in that case
51 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
52 case of MAIL FROM:<> rejection from other early rejections (e.g.
53 rejection of HELO). When verifying a recipient using a non-null MAIL
54 address, the cache is ignored if it shows MAIL FROM:<> rejection.
55 Whatever the result of the callout, the value of the domain cache is
56 left unchanged (for any other kind of callout, getting as far as trying
57 RCPT means that the domain itself is ok).
59 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
62 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
63 manifest itself as EPIPE rather than ECONNECT. When tidying away a
64 session, the daemon ignores ECONNECT errors and logs others; it now
65 ignores EPIPE as well.
67 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
68 (quoted-printable decoding).
70 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
71 later the small subsequent patch to fix an introduced bug.
73 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
75 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
77 PH/08 An error is now given if message_size_limit is specified negative.
79 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
80 to be given (somewhat) arbitrary names.
82 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
85 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
86 miscellaneous code fixes
88 PH/10 Added the log_reject_target ACL modifier to specify where to log
91 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
92 hostname. This is wrong, because it relates to the incoming message (and
93 probably the interface on which it is arriving) and not to the outgoing
94 callout (which could be using a different interface). This has been
95 changed to use the value of the helo_data option from the smtp transport
96 instead - this is what is used when a message is actually being sent. If
97 there is no remote transport (possible with a router that sets up host
98 addresses), $smtp_active_hostname is used.
100 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
101 tweaks were necessary in order to get it to work (see also 21 below):
102 (a) The code assumed that strncpy() returns a negative number on buffer
103 overflow, which isn't the case. Replaced with Exim's string_format()
105 (b) There were several signed/unsigned issues. I just did the minimum
106 hacking in of casts. There is scope for a larger refactoring.
107 (c) The code used strcasecmp() which is not a standard C function.
108 Replaced with Exim's strcmpic() function.
109 (d) The code set only $1; it now sets $auth1 as well.
110 (e) A simple test gave the error "authentication client didn't specify
111 service in request". It would seem that Dovecot has changed its
112 interface. Fortunately there's a specification; I followed it and
113 changed what the client sends and it appears to be working now.
115 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
118 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
119 address A is aliased to B and C, where B exists and C does not. Without
120 -v the output is "A verified" because verification stops after a
121 successful redirection if more than one address is generated. However,
122 with -v the child addresses are also verified. Exim was outputting "A
123 failed to verify" and then showing the successful verification for C,
124 with its parentage. It now outputs "B failed to verify", showing B's
125 parentage before showing the successful verification of C.
127 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
128 look up a TXT record in a specific list after matching in a combined
131 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
132 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
133 they consult the DNS. I had assumed they would set it the way they
134 wanted; and indeed my experiments on Linux seem to show that in some
135 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
136 To be on the safe side, however, I have now made the interface to
137 host_find_byname() similar to host_find_bydns(), with an argument
138 containing the DNS resolver options. The host_find_byname() function now
139 sets these options at its start, just as host_find_bydns() does. The smtp
140 transport options dns_qualify_single and dns_search_parents are passed to
141 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
142 of host_find_byname() use the default settings of RES_DEFNAMES
143 (qualify_single) but not RES_DNSRCH (search_parents).
145 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
146 spool_read_header() do less string testing, by means of a preliminary
147 switch on the second character of optional "-foo" lines. (This is
148 overdue, caused by the large number of possibilities that now exist.
149 Originally there were few.) While I was there, I also converted the
150 str(n)cmp tests so they don't re-test the leading "-" and the first
151 character, in the hope this might squeeze out yet more improvement.
153 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
154 flag allowing group syntax was set by the header_syntax check but not
155 turned off, possible causing trouble later; (2) The flag was not being
156 set at all for the header_verify test, causing "group"-style headers to
157 be rejected. I have now set it in this case, and also caused header_
158 verify to ignore an empty address taken from a group. While doing this, I
159 came across some other cases where the code for allowing group syntax
160 while scanning a header line wasn't quite right (mostly, not resetting
161 the flag correctly in the right place). These bugs could have caused
162 trouble for malformed header lines. I hope it is now all correct.
164 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
165 with the "reply" argument non-NULL. The code, however (which originally
166 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
167 but it didn't always do it. This confused somebody who was copying the
168 code for some other use. I have removed all the tests.
170 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
171 feature that was used to support insecure browsers during the U.S. crypto
172 embargo. It requires special client support, and Exim is probably the
173 only MTA that supported it -- and would never use it because real RSA is
174 always available. This code has been removed, because it had the bad
175 effect of slowing Exim down by computing (never used) parameters for the
176 RSA_EXPORT functionality.
178 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
179 authenticator to fail if there's a tab character in the incoming data
180 (there should never be unless someone is messing about, as it's supposed
181 to be base64-encoded). Also added, on Timo's advice, the "secured" option
182 if the connection is using TLS or if the remote IP is the same as the
183 local IP, and the "valid-client-cert option" if a client certificate has
186 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
187 authenticators. This can be used for authorization after authentication
188 succeeds. (In the case of plaintext, it servers for both authentication
191 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
192 if any retry times were supplied.
194 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
195 connection, where there is no client IP address to check. In this
196 situation, the verify now always succeeds.
198 PH/25 Applied John Jetmore's -Mset patch.
200 PH/26 Added -bem to be like -Mset, but loading a message from a file.
202 PH/27 In a string expansion for a processed (not raw) header when multiple
203 headers of the same name were present, leading whitespace was being
204 removed from all of them, but trailing whitespace was being removed only
205 from the last one. Now trailing whitespace is removed from each header
206 before concatenation. Completely empty headers in a concatenation (as
209 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
210 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
212 PH/29 [Removed. This was a change that I later backed out, and forgot to
213 correct the ChangeLog entry (that I had efficiently created) before
214 committing the later change.]
216 PH/30 Exim was sometimes attempting to deliver messages that had suffered
217 address errors (4xx response to RCPT) over the same connection as other
218 messages routed to the same hosts. Such deliveries are always "forced",
219 so retry times are not inspected. This resulted in far too many retries
220 for the affected addresses. The effect occurred only when there were more
221 hosts than the hosts_max_try setting in the smtp transport when it had
222 the 4xx errors. Those hosts that it had tried were not added to the list
223 of hosts for which the message was waiting, so if all were tried, there
224 was no problem. Two fixes have been applied:
226 (i) If there are any address or message errors in an SMTP delivery, none
227 of the hosts (tried or untried) are now added to the list of hosts
228 for which the message is waiting, so the message should not be a
229 candidate for sending over the same connection that was used for a
230 successful delivery of some other message. This seems entirely
231 reasonable: after all the message is NOT "waiting for some host".
232 This is so "obvious" that I'm not sure why it wasn't done
233 previously. Hope I haven't missed anything, but it can't do any
234 harm, as the worst effect is to miss an optimization.
236 (ii) If, despite (i), such a delivery is accidentally attempted, the
237 routing retry time is respected, so at least it doesn't keep
238 hammering the server.
240 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
241 in ${readsocket because some servers need this prod.
243 PH/32 Added some extra debug output when updating a wait-xxx database.
245 PH/33 The hint "could be header name not terminated by colon", which has been
246 given for certain expansion errors for a long time, was not being given
247 for the ${if def:h_colon_omitted{... case.
249 PH/34 The spec says: "With one important exception, whenever a domain list is
250 being scanned, $domain contains the subject domain." There was at least
251 one case where this was not true.
253 PH/35 The error "getsockname() failed: connection reset by peer" was being
254 written to the panic log as well as the main log, but it isn't really
255 panic-worthy as it just means the connection died rather early on. I have
256 removed the panic log writing for the ECONNRESET error when getsockname()
259 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
260 runs only) independently of the message's sender address. This meant
261 that, if the 4xx error was in fact related to the sender, a different
262 message to the same recipient with a different sender could confuse
263 things. In particualar, this can happen when sending to a greylisting
264 server, but other circumstances could also provoke similar problems.
265 I have changed the default so that the retry time for these errors is now
266 based a combination of the sender and recipient addresses. This change
267 can be overridden by setting address_retry_include_sender=false in the
270 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
271 remote server are returned as part of bounce messages. This was not
272 happening for LMTP over a pipe (the lmtp transport), but now it is the
273 same for both kinds of LMTP.
275 PH/38 Despite being documented as not happening, Exim was rewriting addresses
276 in header lines that were in fact CNAMEs. This is no longer the case.
278 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
279 and queue runs started by the daemon processed all messages. This has
280 been fixed so that -R and -S can now usefully be given with -q<time>.
282 PH/40 Import PCRE release 6.7 (fixes some bugs).
284 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
286 PH/42 Give an error if -q is specified more than once.
288 PH/43 Renamed the variables $interface_address and $interface_port as
289 $received_ip_address and $received_port, to make it clear that these
290 values apply to message reception, and not to the outgoing interface when
291 a message is delivered. (The old names remain recognized, of course.)
293 PH/44 There was no timeout on the connect() call when using a Unix domain
294 socket in the ${readsocket expansion. There now is.
296 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
297 be meaningful with "accept".
299 SC/01 Eximstats V1.43
300 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
302 SC/02 Eximstats V1.44
303 Use a glob alias rather than an array ref in the generated
304 parser. This improves both readability and performance.
306 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
307 Collect SpamAssassin and rejection statistics.
308 Don't display local sender or destination tables unless
309 there is data to show.
310 Added average volumes into the top table text output.
312 SC/04 Eximstats V1.46
313 Collect data on the number of addresses (recipients)
314 as well as the number of messages.
316 SC/05 Eximstats V1.47
317 Added 'Message too big' to the list of mail rejection
318 reasons (thanks to Marco Gaiarin).
320 SC/06 Eximstats V1.48
321 Mainlog lines which have GMT offsets and are too short to
322 have a flag are now skipped.
324 SC/07 Eximstats V1.49 (Alain Williams)
325 Added the -emptyok flag.
327 SC/08 Eximstats V1.50
328 Fixes for obtaining the IP address from reject messages.
330 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
331 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
332 whitesspace changes from 4.64-PH/27
334 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
337 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
338 are found, allow negative numbers in numeric criteria)
340 JJ/06 exipick.20061117.2, added new $message_body_missing variable
342 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
343 to match changes made in 4.64-PH/43
345 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
347 PH/47 Put in an explicit test for a DNS lookup of an address record where the
348 "domain" is actually an IP address, and force a failure. This locks out
349 those revolvers/nameservers that support "A-for-A" lookups, in
350 contravention of the specifications.
352 PH/48 When a host name was looked up from an IP address, and the subsequent
353 forward lookup of the name timed out, the host name was left in
354 $sender_host_name, contrary to the specification.
356 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
357 restricted to single-key lookups, Exim was not diagnosing an error if
358 * or *@ was used with a query-style lookup.
360 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
362 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
363 long ago, but noone interested enough thought of it. Let's just say that
364 the "1.1" means that there are some new functions that weren't there at
365 some point in the past.
367 PH/51 Error processing for expansion failure of helo_data from an smtp
368 transport during callout processing was broken.
370 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
371 tested/used via the -bh/-bhc/-bs options.
373 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
374 bug, fixed in subsequent PCRE releases).
376 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
377 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
379 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
385 SC/01 Use a glob alias rather than an array ref in eximstats generated
386 parser. This improves both readability and performance.
388 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
389 Don't display local sender or destination tables in eximstats unless
390 there is data to show.
391 Added average volumes into the eximstats top table text output.
393 SC/03 Collect data on the number of addresses (recipients) as well
394 as the number of messages in eximstats.
396 TF/01 Correct an error in the documentation for the redirect router. Exim
397 does (usually) call initgroups() when daemonizing.
399 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
400 with consistent privilege compared to when running as a daemon.
402 TF/03 Note in the spec that $authenticated_id is not set for local
403 submissions from trusted users.
405 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
406 Thanks to Dean Brooks <dean@iglou.com> for the patch.
408 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
409 by adding some example configuration directives to the default
410 configuration file. A little bit of work is required to uncomment the
411 directives and define how usernames and passwords are checked, but
412 there is now a framework to start from.
414 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
415 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
416 without this. I don't know how relevant this is to other LDAP libraries.
418 PH/02 Add the verb name to the "unknown ACL verb" error.
420 PH/03 Magnus Holmgren's patch for filter_prepend_home.
422 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
424 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
425 directory not expanded when it should be if an expanded home directory
426 was set for the address (which is overridden by the transport).
428 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
431 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
432 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
433 because it is too late at that time, and has no effect.
435 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
436 security issue with \' (bugzilla #107). I could not use the
437 PQescapeStringConn() function, because it needs a PGconn value as one of
440 PH/08 When testing addresses using -bt, indicate those final addresses that
441 are duplicates that would not cause an additional delivery. At least one
442 person was confused, thinking that -bt output corresponded to deliveries.
443 (Suppressing duplicates isn't a good idea as you lose the information
444 about possibly different redirections that led to the duplicates.)
446 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
447 systems where poll() doesn't work, in particular OS X.
449 PH/10 Added more information to debugging output for retry time not reached.
451 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
452 operations in malware.c.
454 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
457 PH/13 If write_rejectlog was set false when logging was sent to syslog with
458 syslog_duplication set false, log lines that would normally be written
459 both the the main log and to the reject log were not written to syslog at
462 PH/14 In the default configuration, change the use of "message" in ACL warn
463 statements to "add_header".
465 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
466 not followed by a command (e.g. "seen endif").
468 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
469 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
472 PH/17 Added extra conditions to the default value of delay_warning_condition
476 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
477 { match{$h_precedence:}{(?i)bulk|list|junk} } \
478 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
481 The Auto-Submitted: and various List- headers are standardised, whereas I
482 don't think Precedence: ever was.
484 PH/18 Refactored debugging code in route_finduser() to show more information,
485 in particular, the error code if getpwnam() issues one.
487 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
488 This is apparently needed in addition to the PH/07 change above to avoid
489 any possible encoding problems.
491 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
492 but not after initializing Perl.
494 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
495 output them only if debugging. By default they are written stderr,
496 apparently, which is not desirable.
498 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
501 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
504 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
506 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
507 authenticated or an ident call has been made. Suppress the default
508 values for $authenticated_id and $authenticated_sender (but permit -oMai
509 and -oMas) when testing with -bh.
511 PH/24 Re-jigged the order of the tests in the default configuration so that the
512 tests for valid domains and recipients precede the DNS black list and CSA
513 tests, on the grounds that those ones are more expensive.
515 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
516 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
517 This bug exists in every version of Exim that I still have, right back to
520 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
521 However, an attempt to turn on case-sensitivity in a regex key by
522 including (?-i) didn't work because the subject string was already
523 lowercased, and the effects were non-intuitive. It turns out that a
524 one-line patch can be used to allow (?-i) to work as expected.
530 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
531 other effects) broke the use of negated acl sub-conditions.
533 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
536 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
537 "Deny" causes Exim to reject the incoming connection with a 554 error.
538 Unfortunately, if there is a major crisis, such as a disk failure,
539 tcp-wrappers gives "deny", whereas what one would like would be some
540 kind of temporary error. A kludge has been added to help with this.
541 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
542 554 error is used if errno is still zero or contains ENOENT (which occurs
543 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
546 PH/03 Add -lutil to the default FreeBSD LIBS setting.
548 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
549 errors. Otherwise a message that provokes a temporary error (when other
550 messages do not) can cause a whole host to time out.
552 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
553 addresses were routed directly to files or pipes from a redirect router.
554 File deliveries just didn't batch; pipe deliveries might have suffered
557 PH/06 A failure to get a lock for a hints database would erroneously always say
558 "Failed to get write lock", even when it was really a read lock.
560 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
561 of 0600. This has been changed to use the value of the lockfile_mode
562 option (which defaults to 0600).
564 PH/08 Applied small patch from the Sieve maintainer.
566 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
567 folder from quota calculations, a direct delivery into this folder messed
568 up the contents of the maildirsize file. This was because the regex was
569 used only to exclude .Trash (or whatever) when the size of the mailbox
570 was calculated. There was no check that a delivery was happening into an
571 excluded directory. This bug has been fixed by ignoring all quota
572 processing for deliveries into excluded directories.
574 PH/10 Added the maildirfolder_create_regex option to appendfile.
580 PH/01 The code for finding all the local interface addresses on a FreeBSD
581 system running IPv6 was broken. This may well have applied to all BSD
582 systems, as well as to others that have similar system calls. The broken
583 code found IPv4 interfaces correctly, but gave incorrect values for the
584 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
585 that it would not match correctly against @[] and not recognize the IPv6
588 PH/02 The ipliteral router was not recognizing addresses of the form user@
589 [ipv6:....] because it didn't know about the "ipv6:" prefix.
591 PH/03 Added disable_ipv6.
593 PH/04 Changed $reply_address to use the raw form of the headers instead of the
594 decoded form, because it is most often used to construct To: headers
595 lines in autoreplies, and the decoded form may well be syntactically
596 invalid. However, $reply_address has leading white space removed, and all
597 newlines turned into spaces so that the autoreply transport does not
600 PH/05 If group was specified without a user on a router, and no group or user
601 was specified on a transport, the group from the router was ignored.
603 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
604 for visible compile-time settings that can be used to change these
605 numbers, for those that want even more. Backwards compatibility with old
606 spool files has been maintained. However, going back to a previous Exim
607 release will lost any variables that are in spool files.
609 PH/07 Two small changes when running in the test harness: increase delay when
610 passing a TCP/IP connection to a new process, in case the original
611 process has to generate a bounce, and remove special handling of
612 127.0.0.2 (sic), which is no longer necessary.
614 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
615 be the same on different OS.
617 PH/09 Moved a debug statement in filter processing to avoid a race problem when
620 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
621 whether --show-vars was specified or not
623 JJ/02 exipick: Added support for new ACL variable spool format introduced
626 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
627 syntactically invalid From: or Reply-to: line, and a filter used this to
628 generate an autoreply, and therefore failed to obtain an address for the
629 autoreply, Exim could try to deliver to a non-existent relative file
630 name, causing unrelated and misleading errors. What now happens is that
631 it logs this as a hard delivery error, but does not attempt to create a
634 PH/11 The exinext utility has a -C option for testing purposes, but although
635 the given file was scanned by exinext itself; it wasn't being passed on
636 when Exim was called.
638 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
639 an end-of-file indication when reading a command response.
641 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
642 compiled. In many other places in Exim, IPv6 addresses are always
643 recognized, so I have changed this. It also means that IPv4 domain
644 literals of the form [IPV4:n.n.n.n] are now always recognized.
646 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
647 used if the router is not running as root, for example, when verifying at
648 ACL time, or when using -bh. The debugging output from this situation was
649 non-existent - all you got was a failure to exec. I have made two
652 (a) Failures to set uid/gid, the current directory, or a process leader
653 in a subprocess such as that created by queryprogram now generate
654 suitable debugging ouput when -d is set.
656 (b) The queryprogram router detects when it is not running as root,
657 outputs suitable debugging information if -d is set, and then runs
658 the subprocess without attempting to change uid/gid.
660 PH/15 Minor change to Makefile for building test_host (undocumented testing
663 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
664 additional section of a DNS packet that returns MX or SRV records.
665 Instead, it always explicitly searches for A/AAAA records. This avoids
666 major problems that occur when a DNS server includes only records of one
667 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
668 fixed another bug: if SRV records were looked up and the corresponding
669 address records were *not* found in the additional section, the port
670 values from the SRV records were lost.
672 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
673 using the correct key (the original address) when searching the retry
674 rules in order to find which one to use for generating the retry hint.
676 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
677 adding the default one. Similarly, if it contains a Reply-To: header, the
678 errors_reply_to option, if set, is not used.
680 PH/19 When calculating a retry time, Exim used to measure the "time since
681 failure" by looking at the "first failed" field in the retry record. Now
682 it does not use this if it is later than than the arrival time of the
683 message. Instead it uses the arrival time. This makes for better
684 behaviour in cases where some deliveries succeed, thus re-setting the
685 "first failed" field. An example is a quota failure for a huge message
686 when small messages continue to be delivered. Without this change, the
687 "time since failure" will always be short, possible causing more frequent
688 delivery attempts for the huge message than are intended.
689 [Note: This change was subsequently modified - see PH/04 for 4.62.]
691 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
692 $1, $2, $3) because the numerical variables can be reset during some
693 expansion items (e.g. "match"), thereby losing the authentication data.
695 PH/21 Make -bV show the size of off_t variables so that the test suite can
696 decide whether to run tests for quotas > 2G.
698 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
699 mailbox_size, and mailbox_filecount in the appendfile transport. If a
700 filecount value is greater than 2G or if a quota value is greater than 2G
701 on a system where the size of off_t is not greater than 4, a panic error
704 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
705 never match. The debug and -bh output now contains an explicit error
706 message indicating a malformed IPv4 address or mask.
708 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
709 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
712 PH/25 Do not write to syslog when running in the test harness. The only
713 occasion when this arises is a failure to open the main or panic logs
714 (for which there is an explicit test).
716 PH/26 Added the /no_tell option to "control=freeze".
718 PH/27 If a host name lookup failed very early in a connection, for example, if
719 the IP address matched host_lookup and the reverse lookup yielded a name
720 that did not have a forward lookup, an error message of the form "no IP
721 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
722 could be logged. Now it outputs the IP address instead of "NULL".
724 PH/28 An enabling patch from MH: add new function child_open_exim2() which
725 allows the sender and the authenticated sender to be set when
726 submitting a message from within Exim. Since child_open_exim() is
727 documented for local_scan(), the new function should be too.
729 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
730 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
731 results in an empty string is now treated as unset.
733 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
735 PH/31 Added sender_verify_fail logging option.
737 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
738 needed by RFC 822 but not by RFC 2822 was commented out. I have now
739 tidied the source and removed it altogether.
741 PH/33 When a queue run was abandoned because the load average was too high, a
742 log line was always written; now it is written only if the queue_run log
743 selector is set. In addition, the log line for abandonment now contains
744 information about the queue run such as the pid. This is always present
745 in "start" and "stop" lines but was omitted from the "abandon" line.
747 PH/34 Omit spaces between a header name and the colon in the error message that
748 is given when verify = headers_syntax fails (if there are lots of them,
749 the message gets confusing).
751 PH/35 Change the default for dns_check_names_pattern to allow slashes within
752 names, as there are now some PTR records that contain slashes. This check
753 is only to protect against broken name servers that fall over on strange
754 characters, so the fact that it applies to all lookups doesn't matter.
756 PH/36 Now that the new test suite is complete, we can remove some of the
757 special code in Exim that was needed for the old test suite. For example,
758 sorting DNS records because real resolvers return them in an arbitrary
759 order. The new test suite's fake resolver always returns records in the
762 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
763 bounces) except when queue_only is set, to avoid logging races between
764 the different processes.
766 PH/38 Panic-die if .include specifies a non-absolute path.
768 PH/39 A tweak to the "H" retry rule from its user.
770 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
771 a label. They prevented compilation on older perls.
773 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
774 a warning to be raised on newish perls.
776 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
777 on queue. Changes to match documented behaviour of showing count of
778 messages matching specified criteria.
780 PH/40 Changed the default ident timeout from 30s to 5s.
782 PH/41 Added support for the use of login_cap features, on those BSD systems
783 that have them, for controlling the resources used by pipe deliveries.
785 PH/42 The content-scanning code uses fopen() to create files in which to put
786 message data. Previously it was not paying any attention to the mode of
787 the files. Exim runs with umask(0) because the rest of the code creates
788 files with open(), and sets the required mode explicitly. Thus, these
789 files were ending up world-writeable. This was not a big issue, because,
790 being within the spool directory, they were not world-accessible. I have
791 created a function called modefopen, which takes an additional mode
792 argument. It sets umask(777), creates the file, chmods it to the required
793 mode, then resets the umask. All the relevant calls to fopen() in the
794 content scanning code have been changed to use this function.
796 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
797 to 24 hours. This avoids potential overflow problems when processing G
798 and H retry rules. I suspect nobody ever tinkers with this value.
800 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
802 PH/45 When the plaintext authenticator is running as a client, the server's
803 challenges are checked to ensure they are valid base64 strings. By
804 default, the authentication attempt is cancelled if an invalid string is
805 received. Setting client_ignore_invalid_base64 true ignores these errors.
806 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
807 they are received. Thus, the responses can be made to depend on the
808 challenges. If an invalid string is ignored, an empty string is placed in
811 PH/46 Messages that are created by the autoreply transport now contains a
812 References: header, in accordance with RFCs 2822 and 3834.
814 PH/47 Added authenticated_sender_force to the smtp transport.
816 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
818 PH/49 Installed latest patch from the Sieve maintainer.
820 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
821 was also set, the appendfile transport was unnecessarily scanning a
822 directory of message files (e.g. for maildir delivery) to find the count
823 of files (along with the size), even though it did not need this
824 information. It now does the scan only if it needs to find either the
825 size of the count of files.
827 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
829 PH/52 Two bugs concerned with error handling when the smtp transport is
832 (i) Exim was not creating retry information for temporary errors given
833 for individual recipients after the DATA command when the smtp transport
834 was used in LMTP mode. This meant that they could be retried too
835 frequently, and not timed out correctly.
837 (ii) Exim was setting the flag that allows error details to be returned
838 for LMTP errors on RCPT commands, but not for LMTP errors for individual
839 recipients that were returned after the DATA command.
841 PH/53 This is related to PH/52, but is more general: for any failing address,
842 when detailed error information was permitted to be returned to the
843 sender, but the error was temporary, then after the final timeout, only
844 "retry timeout exceeded" was returned. Now it returns the full error as
845 well as "retry timeout exceeded".
847 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
848 do this, and (what is worse) MTAs that accept it.
850 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
851 will now be deprecated.
853 PH/56 New os.c-cygwin from the Cygwin maintainer.
855 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
856 formats (previously only available in exim formats via -bpr, -bpru,
857 and -bpra. Now also available in native and exiqgrep formats)
859 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
860 with very large, slow to parse queues
862 JJ/08 exipick: added ! as generic prefix to negate any criteria format
864 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
866 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
867 responses to authentication challenges, though it was showing the
868 challenges; (ii) I've removed the CR characters from the debug output for
871 PH/58 Allow for the insertion of a newline as well as a space when a string
872 is turned into more than one encoded-word during RFC 2047 encoding. The
873 Sieve code now uses this.
875 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
876 data_4xx, lost_connection, tls_required.
878 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
879 message was being sent as an SMTP response.
881 PH/61 Add -l and -k options to exicyclog.
883 PH/62 When verifying, if an address was redirected to one new address, so that
884 verification continued, and the new address failed or deferred after
885 having set something in $address_data, the value of $address_data was not
886 passed back to the ACL. This was different to the case when no
887 redirection occurred. The value is now passed back in both cases.
889 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
890 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
891 use login_cap.h, so on its own it isn't the distinguishing feature. The
892 new name refers directly to the setclassresources() function.
894 PH/65 Added configuration files for NetBSD3.
896 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
898 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
899 is preferred over IPv4.
901 PH/68 The bounce_return_message and bounce_return_body options were not being
902 honoured for bounces generated during the reception of non-SMTP messages.
903 In particular, this applied to messages rejected by the ACL. This bug has
904 been fixed. However, if bounce_return_message is true and bounce_return_
905 body is false, the headers that are returned for a non-SMTP message
906 include only those that have been read before the error was detected.
907 (In the case of an ACL rejection, they have all been read.)
909 PH/69 The HTML version of the specification is now built in a directory called
910 spec_html instead of spec.html, because the latter looks like a path with
911 a MIME-type, and this confuses some software.
913 PH/70 Catch two compiler warnings in sieve.c.
915 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
916 function verify_get_ident() calls ip_connect() to connect a socket, but
917 if the "connect()" function timed out, ip_connect() used to close the
918 socket. However, verify_get_ident() also closes the socket later, and in
919 between Exim writes to the log, which may get opened at this point. When
920 the socket was closed in ip_connect(), the log could get the same file
921 descriptor number as the socket. This naturally causes chaos. The fix is
922 not to close the socket in ip_connect(); the socket should be closed by
923 the function that creates it. There was only one place in the code where
924 this was missing, in the iplookup router, which I don't think anybody now
925 uses, but I've fixed it anyway.
927 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
928 well as to direct DNS lookups. Otherwise the handling of names in host
929 lists is inconsistent and therefore confusing.
935 PH/01 Two changes to the default runtime configuration:
937 (1) Move the checks for relay_from_hosts and authenticated clients from
938 after to before the (commented out) DNS black list checks.
940 (2) Add control=submission to the relay_from_hosts and authenticated
941 clients checks, on the grounds that messages accepted by these
942 statements are most likely to be submissions.
944 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
946 (1) Generate an error if the third argument for the ${prvs expansion is
949 (2) Treat a missing third argument of ${prvscheck as if it were an empty
952 (3) Reset the variables that are obtained from the first argument of
953 ${prvscheck and used in the second argument before leaving the code,
954 because their memory is reclaimed, so using them afterwards may do
957 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
958 one (it's much easier than Tom thought :-).
960 (5) Because of (4), we can now allow for the use of $prvscheck_result
961 inside the third argument.
963 PH/03 For some reason, the default setting of PATH when running a command from
964 a pipe transport was just "/usr/bin". I have changed it to
967 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
968 anything to be listed in the output from -bV.
970 PH/05 When a filter generated an autoreply, the entire To: header line was
971 quoted in the delivery log line, like this:
973 => >A.N.Other <ano@some.domain> <original@ddress> ...
975 This has been changed so that it extracts the operative address. There
976 may be more than one such address. If so, they are comma-separated, like
979 => >ano@some.domain,ona@other.domain <original@ddress> ...
981 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
982 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
983 not being looked up in the rDNS to get a host name, Exim was showing the
984 IP address twice in Received: lines, even though the IP addresses were
985 identical. For example:
987 Received: from [1.2.3.4] (helo=[1.2.3.4])
989 However, if the real host name was known, it was omitting the HELO data
990 if it matched the actual IP address. This has been tidied up so that it
991 doesn't show the same IP address twice.
993 PH/07 When both +timestamp and +memory debugging was on, the value given by
994 $tod_xxx expansions could be wrong, because the tod_stamp() function was
995 called by the debug printing, thereby overwriting the timestamp buffer.
996 Debugging no longer uses the tod_stamp() function when +timestamp is set.
998 PH/08 When the original message was included in an autoreply transport, it
999 always said "this is a copy of the message, including all the headers",
1000 even if body_only or headers_only was set. It now gives an appropriate
1003 PH/09 Applied a patch from the Sieve maintainer which:
1005 o fixes some comments
1006 o adds the (disabled) notify extension core
1007 o adds some debug output for the result of if/elsif tests
1008 o points to the current vacation draft in the documentation
1009 and documents the missing references header update
1013 o fixes a bug in processing the envelope test (when testing
1014 multiple envelope elements, the last element determinted the
1017 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1018 Electronic Mail") by including:
1020 Auto-submitted: auto-generated
1022 in the messages that it generates (bounce messages and others, such as
1023 warnings). In the case of bounce messages for non-SMTP mesages, there was
1024 also a typo: it was using "Auto_submitted" (underscore instead of
1025 hyphen). Since every message generated by Exim is necessarily in response
1026 to another message, thes have all been changed to:
1028 Auto-Submitted: auto-replied
1030 in accordance with these statements in the RFC:
1032 The auto-replied keyword:
1034 - SHOULD be used on messages sent in direct response to another
1035 message by an automatic process,
1037 - MUST NOT be used on manually-generated messages,
1039 - MAY be used on Delivery Status Notifications (DSNs) and Message
1040 Disposition Notifications (MDNs),
1042 - MUST NOT be used on messages generated by automatic or periodic
1043 processes, except for messages which are automatic responses to
1046 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1047 to the default Received: header definition.
1049 PH/12 Added log selector acl_warn_skipped (default on).
1051 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1052 variables because (a) they are in the wrong storage pool and (b) even if
1053 they were copied, it wouldn't work properly because of the caching.
1055 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1056 checking when decoding. Apparently there are clients that generate
1057 overlong encoded strings. Why am I not surprised?
1059 PH/15 If the first argument of "${if match_address" was not empty, but did not
1060 contain an "@" character, Exim crashed. Now it writes a panic log message
1061 and treats the condition as false.
1063 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1065 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1066 extension "envelope-auth". The code is finished and in agreement with
1067 other implementations, but there is no documentation so far and in fact,
1068 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1069 not changing the active code.
1071 Print executed "if" and "elsif" statements when debugging is used. This
1072 helps a great deal to understand what a filter does.
1074 Document more things not specified clearly in RFC3028. I had all this
1075 sorted out, when out of a sudden new issues came to my mind. Oops."
1077 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1080 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1081 canonical form (as documented). However, after a host name lookup from
1082 the IP address, check_host() was doing a simple string comparison with
1083 addresses acquired from the DNS when checking that the found name did
1084 have the original IP as one of its addresses. Since any found IPv6
1085 addresses are likely to be in abbreviated form, the comparison could
1086 fail. Luckily, there already exists a function for doing the comparison
1087 by converting both addresses to binary, so now that is used instead of
1088 the text comparison.
1090 PH/20 There was another similar case to PH/19, when a complete host name was
1091 given in a host list; looking up its IP address could give an abbreviated
1092 form, whereas the current host's name might or might not be abbreviated.
1093 The same fix has been applied.
1099 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1100 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1103 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1104 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1106 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1108 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1109 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1110 Also, for "Auto-Submitted", treat anything other than "no" as
1111 non-personal, in accordance with RFC 3834. (Previously it treated
1112 anything starting "auto-" as non-personal.)
1114 TF/01 The control=submission/name=... option had a problem with syntax
1115 errors if the name included a slash character. The /name= option
1116 now slurps the rest of the string, so it can include any characters
1117 but it must come last in the list of options (after /sender_retain
1120 PH/05 Some modifications to the interface to the fake nameserver for the new
1128 TK/01 Added the "success_on_redirect" address verification option. See
1129 NewStuff for rationale and an example.
1131 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1133 PH/02 Patch to exigrep to allow it to work on syslog lines.
1135 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1136 fread() to skip over the body file's header line, because in Cygwin the
1137 header line is locked and is inaccessible.
1139 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1140 co-exist for some time) to make it clear that it is the Exim ID that is
1141 referenced, not the Message-ID: header line.
1143 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1144 string_format() function, because snprintf() does not exist on all
1147 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1148 ${stat: expansion item.
1150 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1151 protocol synchronization error", to keep the pedants happy.
1153 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1154 well as for IRIX systems, when gcc is being used. See the host.c source
1157 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1159 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1162 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1163 transport and to the smtp transport in LMTP mode.
1165 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1167 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1168 run a filter in a subprocess. This could lead to confusion in subsequent
1169 lookups in the parent process. There should also be a search_tidyup() at
1170 the end of the subprocess.
1172 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1173 only if the host matched helo_try_verify_hosts, which caused the
1174 verification to occur when the EHLO/HELO command was issued. The ACL just
1175 tested the remembered result. Now, if a previous verification attempt has
1176 not happened, "verify = helo" does it there and then.
1178 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1180 TK/03 Fix log output including CR from clamd.
1182 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1183 exist provoked a memory error which could cause a segfault.
1185 PH/15 Installed PCRE 6.2
1187 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1189 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1190 of the problem. Specifically, suggested +O2 rather than +O1 for the
1193 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1195 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1196 record, the port was not correctly logged when the outgoing_port log
1197 selector was set (it logged the transort's default port).
1199 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1200 fallback_hosts, and "hosts" in the smtp transport.
1202 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1203 host errors such as "Connection refused".
1205 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1206 authentication with radiusclient 0.4.9:
1208 - Error returned from rc_read_config was caught wrongly
1209 - Username/password not passed on to radius server due to wrong length.
1211 The presumption is that some radiusclient API changes for 4.51/PH/17
1212 were not taken care of correctly. The code is still untested by me (my
1213 Linux distribution still has 0.3.2 of radiusclient), but it was
1214 contributed by a Radius user.
1216 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1217 expanding the "port" option of the smtp transport.
1219 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1220 while reading a MIME header. Thanks to Tom Hughes for a patch.
1222 PH/24 Include config.h inside local_scan.h so that configuration settings are
1225 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1226 that in an ACL for RCPT (for example), you can examine exactly what was
1229 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1230 commands, but it was not correctly comparing the address with the actual
1231 client host address. Thus, it would show the EHLO address in Received:
1232 header lines when this was not necessary.
1234 PH/27 Added the % operator to ${eval:}.
1236 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1237 it should be ignoring failures (because with -C, for example, it has lost
1238 privilege). It wasn't ignoring creation failures other than "already
1241 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1244 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1245 because an input line was too long, either on its own, or by virtue of
1246 too many continuations, the temporary file was not being removed, and the
1247 return code was incorrect.
1249 PH/31 Missing "BOOL" in function definition in filtertest.c.
1251 PH/32 Applied Sieve patches from the maintainer.
1253 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1255 PH/33 Added "verify = not_blind".
1257 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1258 Local/Makefile (with some defaults set). These are used in built scripts
1259 such as exicyclog, but they have never been used in the exim_install
1260 script (though there are many overriding facilities there). I have
1261 arranged that the exim_install script now takes note of these two
1264 PH/35 Installed configuration files for Dragonfly.
1266 PH/36 When a locally submitted message by a trusted user did not contain a
1267 From: header, and the sender address was obtained from -f or from an SMTP
1268 MAIL command, and the trusted user did not use -F to supply a sender
1269 name, $originator_name was incorrectly used when constructing a From:
1270 header. Furthermore, $originator_name was used for submission mode
1271 messages from external hosts without From: headers in a similar way,
1272 which is clearly wrong.
1274 PH/37 Added control=suppress_local_fixups.
1276 PH/38 When log_selector = +received_sender was set, and the addition of the
1277 sender made the log line's construction buffer exactly full, or one byte
1278 less than full, an overflow happened when the terminating "\n" was
1281 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1282 when the result of a list match is failure because a DNS lookup failed.
1284 PH/40 RM_COMMAND is now used in the building process.
1286 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1287 the "build-* directories that it finds.
1289 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1290 address in a domain literal was a prefix of an interface address.
1292 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1293 when verifying a sender address, unless rewrite_headers is false.
1295 PH/44 Wrote a long comment about why errors_to addresses are verified as
1296 recipients, not senders.
1298 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1299 the ratelimit ACL was added.
1301 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1303 PH/46 Added extra information about PostgreSQL errors to the error string.
1305 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1306 avoiding the need to install special zones in a real server. This is
1307 backwards compatible; if it can't find the fake resolver, it drops back.
1308 Thus, both old and new test suites can be run.
1310 TF/02 Added util/ratelimit.pl
1312 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1313 clock is set back in time.
1315 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1316 Candler <B.Candler@pobox.com>.
1318 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1319 OK for addresses that are the result of redirections.
1321 PH/48 A number of further additions for the benefit of the new test suite,
1322 including a fake gethostbyname() that interfaces to the fake DNS resolver
1325 TF/06 The fix for widen_domains has also been applied to qualify_single and
1326 search_parents which are the other dnslookup options that can cause
1329 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1332 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1334 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1335 service" if it encountered an MX record with an empty target hostname.
1336 The message is now "an MX or SRV record indicated no SMTP service".
1338 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1339 if the DNS of the sending site is misconfigured. This is quite a
1340 common situation. This change restores the behaviour of treating a
1341 helo verification defer as a failure.
1343 PH/51 If self=fail was set on a router, the bounce message did not include the
1344 actual error message.
1350 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1352 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1353 command itself ended in error, the underlying message about the transport
1354 filter timeout was being overwritten with the pipe command error. Now the
1355 underlying error message should be appended to the second error message.
1357 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1359 PH/02 Reduce the amount of output that "make" produces by default. Full output
1360 can still be requested.
1362 PH/03 The warning log line about a condition test deferring for a "warn" verb
1363 was being output only once per connection, rather than after each
1364 occurrence (because it was using the same function as for successful
1365 "warn" verbs). This seems wrong, so I have changed it.
1367 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1368 it should not have, which might have caused a crash in the right
1369 circumstances, but probably never did.
1371 PH/04 Installed a modified version of Tony Finch's patch to make submission
1372 mode fix the return path as well as the Sender: header line, and to
1373 add a /name= option so that you can make the user's friendly name appear
1376 TF/03 Added the control = fakedefer ACL modifier.
1378 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1379 Mark Lowes for thorough testing.
1381 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1383 TK/03 Merged latest SRS patch from Miles Wilton.
1385 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1386 with the definition in sysexits.h (which is #included earlier).
1387 Fortunately, Exim does not actually use EX_OK. The code used to try to
1388 preserve the sysexits.h value, by assumimg that macro definitions were
1389 scanned for macro replacements. I have been disabused of this notion,
1390 so now the code just undefines EX_OK before #including unistd.h.
1392 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1393 in the smtp transport. When a block could not be written in a single
1394 write() function, the timeout was being re-applied to each part-write.
1395 This seems wrong - if the receiver was accepting one byte at a time it
1396 would take for ever. The timeout is now adjusted when this happens. It
1397 doesn't have to be particularly precise.
1399 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1400 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1402 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1403 without a domain if the check to <postmaster@domain> fails.
1405 SC/01 Eximstats: added -xls and the ability to specify output files
1406 (patch written by Frank Heydlauf).
1408 SC/02 Eximstats: use FileHandles for outputing results.
1410 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1412 SC/04 Eximstats: fixed display of large numbers with -nvr option
1414 SC/05 Eximstats: fixed merging of reports with empty tables.
1416 SC/06 Eximstats: added the -include_original_destination flag
1418 SC/07 Eximstats: removed tabs and trailing whitespace.
1420 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1422 TK/06 MBOX spool code: Add real "From " MBOX separator line
1423 so the .eml file is really in mbox format (even though
1424 most programs do not really care). Patch from Alex Miller.
1426 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1427 The latter is generated from $received_to and is only set if the
1428 message has one envelope recipient. SA can use these headers,
1429 obviously out-of-the-box. Patch from Alex Miller.
1431 PH/08 The ${def test on a variable was returning false if the variable's
1432 value was "0", contrary to what the specification has always said!
1433 The result should be true unless the variable is empty.
1435 PH/09 The syntax error of a character other than { following "${if
1436 def:variable_name" (after optional whitespace) was not being diagnosed.
1437 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1438 accidental colon was present, for example, could give incorrect results.
1440 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1441 result is used (not including appendfile, where other changes are about
1444 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1445 This involved changing a lot of size variables from int to off_t. It
1446 should work with maildirs and everything.
1448 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1449 spamd dying while we are connected to it.
1451 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1454 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1455 function for BDB 4.3.
1457 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1459 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1460 the sources for PCRE have been re-arranged and more files are now
1463 PH/15 The code I had for printing potentially long long variables in PH/11
1464 above was not the best (it lost precision). The length of off_t variables
1465 is now inspected at build time, and an appropriate printing format (%ld
1466 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1467 to be "long long int" or "long int". This is needed for the internal
1468 formatting function string_vformat().
1470 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1471 the configuration file to be ":syslog", then the script "guesses" where
1472 the logs files are, rather than using the compiled in default. In our
1473 case the guess is not the same as the compiled default, so the script
1474 suddenly stopped working when I started to use syslog. The patch checks
1475 to see if log_file_path is "". If so, it attempts to read it from exim
1476 with no configuration file to get the compiled in version, before it
1477 falls back to the previous guessing code."
1479 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1480 implementing BATV in an Exim configuration. See NewStuff for the gory
1483 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1484 Makefile that are specific to HP-UX.
1486 PH/18 If the "use_postmaster" option was set for a recipient callout together
1487 with the "random" option, the postmaster address was used as the MAIL
1488 FROM address for the random test, but not for the subsequent recipient
1489 test. It is now used for both.
1491 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1492 patch removes a few documentation additions to RFC 3028, because the
1493 latest draft now contains them. It adds the new en;ascii-case comparator
1494 and a new error check for 8bit text in MIME parts. Comparator and
1495 require names are now matched exactly. I enabled the subaddress
1496 extension, but it is not well tested yet (read: it works for me)."
1498 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1499 rework some of the code of TK/09 above to avoid the hardwired use of
1500 "%lld" and "long long". Replaced the call to snprintf() with a call to
1503 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1504 records point to non-existent hosts", "retry timeout exceeded", and
1505 "retry time not reached for any host after a long failure period".
1507 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1508 experimental DomainKeys support:
1510 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1511 (2) On an error such as an illegally used "control", the wrong name for
1512 the control was given.
1514 These problems did NOT occur unless DomainKeys support was compiled.
1516 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1518 PH/24 Added ${if match_ip condition.
1520 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1521 databases so that it will be absolutely obvious if a crash occurs in the
1522 DB library. This is a regular occurrence (often caused by mis-matched
1525 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1526 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1527 picked up on a user's system that detects such things. There doesn't seem
1528 to be a gcc warning option for this - only an attribute that has to be
1529 put on the function's prototype. It seems that in Fedora Core 4 they have
1530 set this on a number of new functions. No doubt there will be more in due
1533 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1534 specify a transport. However, if an address that was verified by such a
1535 router was the subject of a callout, Exim crashed because it tried to
1536 read the rcpt_include_affixes from the non-existent transport. Now it
1537 just assumes that the setting of that option is false. This bug was
1538 introduced by 4.51/PH/31.
1540 PH/28 Changed -d+all to exclude +memory, because that information is very
1541 rarely of interest, but it makes the output a lot bigger. People tend to
1542 do -d+all out of habit.
1544 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1545 code in os-type was giving problems when libc.so lives in lib64, like on
1548 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1549 aren't the modern standard, and it seems that some systems' include files
1550 don't always have them. Exim was already checking for some of the newer
1551 ones like T_AAAA, and defining it itself. I've added checks for all the
1552 record types that Exim uses.
1554 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1555 not automatically generating a new one, as it is supposed to. This
1556 prevented TLS from working. If the file did exist, but contained invalid
1557 data, a new version was generated, as expected. It was only the case of a
1558 non-existent file that was broken.
1560 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1561 with a change in libdomainkeys > 0.64.
1563 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1564 from DNS. If the selector record carries the flag, it now has
1565 precedence over the domain-wide flag.
1567 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1569 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1570 the use of a port name that isn't defined in /etc/services) occurred, the
1571 message was deferred as in a normal delivery, and thus remained on the
1572 spool, instead of being failed because of the mua_wrapper setting. This
1573 is now fixed, and I tidied up some of the mua_wrapper messages at the
1576 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1577 the messages in a hash of arrays rather than using individual hashes.
1578 This is a bit cleaner and results in dramatic memory savings, albeit
1579 at a slight CPU cost.
1581 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1582 as requested by Marc Sherman.
1584 SC/10 Eximstats: added histograms for user specified patterns as requested
1587 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1589 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1590 fopen() in the content-scanning modules that did not already have it.
1596 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1597 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1599 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1601 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1603 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1604 to test Sieve filters that use "vacation".
1606 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1607 that changes the way the GnuTLS parameters are stored in the cache file.
1608 The new format can be generated externally. For backward compatibility,
1609 if the data in the cache doesn't make sense, Exim assumes it has read an
1610 old-format file, and it generates new data and writes a new file. This
1611 means that you can't go back to an older release without removing the
1614 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1615 work if there are any delivery delays because "one_time" forces the
1616 parent to be marked "delivered", so its unseen clone is never tried
1617 again. For this reason, Exim now forbids the simultaneous setting of
1620 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1621 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1622 entry if you want to know the details. The fix, however, neglected to
1623 consider the case when local delivery batching is involved. The test for
1624 "previously delivered" was not happening when checking to see if an
1625 address could be batched with a previous (undelivered) one; under
1626 certain circumstances this could lead to multiple deliveries to the same
1629 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1630 in its include files, and this causes problems building Exim.
1632 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1633 header_syntax) but Exim was just ignoring anything given after a slash.
1634 In particular, this caused confusion with an attempt to use "verify =
1635 reverse_host_lookup/defer_ok". An error is now given when options are
1636 supplied for verify items that do not have them. (Maybe reverse_host_
1637 lookup should have a defer_ok option, but that's a different point.)
1639 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1640 defined by RFC 821) to 2048, because there were problems with some AUTH
1641 commands, and RFC 1869 says the size should be increased for extended
1642 SMTP commands that take arguments.
1644 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1647 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1648 "unknown" error; now it says that the functionality isn't in the binary.
1650 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1651 an address' error message when a string expansion fails (syntax or
1652 whatever). Otherwise the password may appear in the log. Following change
1653 PH/42 below, there is no longer a chance of it appearing in a bounce
1656 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1658 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1659 of its addresses was ever tried. (Bugzilla bug #2.)
1661 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1662 the result incorrectly in the debug output. (It correctly added a newline
1663 to what was transported.)
1665 TF/01 Added $received_time.
1667 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1668 commented out examples of how to interface to a virus scanner and to
1669 SpamAssassin. Also added commented examples of av_scanner and
1670 spamd_address settings.
1672 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1673 and controls are allowed in which ACLs. There were a couple of minor
1674 errors. Some of the entries in the conditions table (which is a table of
1675 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1676 negation of where the condition IS allowed.
1678 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1680 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1681 header file does not have a version number, so I've had to invent a new
1682 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1683 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1684 radiusclient), but it was contributed by a Radius user.
1686 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1687 files or directories, for OpenSSL.
1689 PH/19 When an Exim process that is running as root has to create an Exim log
1690 file, it does so in a subprocess that runs as exim:exim so as to get the
1691 ownership right at creation (otherwise, other Exim processes might see
1692 the file with the wrong ownership). There was no test for failure of this
1693 fork() call, which would lead to the process getting stuck as it waited
1694 for a non-existent subprocess. Forks do occasionally fail when resources
1695 run out. I reviewed all the other calls to fork(); they all seem to check
1698 PH/20 When checking for unexpected SMTP input at connect time (before writing
1699 the banner), Exim was not dealing correctly with a non-positive return
1700 from the read() function. If the client had disconnected by this time,
1701 the result was a log entry for a synchronization error with an empty
1702 string after "input=" when read() returned zero. If read() returned -1
1703 (an event I could not check), uninitialized data bytes were printed.
1704 There were reports of junk text (parts of files, etc) appearing after
1707 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1709 PH/22 Added support for macro redefinition, and (re)definition in between
1710 driver and ACL definitions.
1712 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1713 forgetting to use the resulting value; it was using the unexpanded value.
1715 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1716 hadn't been configured. The fix is from Juergen Kreileder, who
1717 understands it better than I do:
1719 "Here's what I see happening with three configured cyrus_sasl
1720 authenticators configured (plain, login, cram-md5):
1722 On startup auth_cyrus_sasl_init() gets called for each of these.
1723 This means three calls to sasl_listmech() without a specified mech_list.
1724 => SASL tests which mechs of all available mechs actually work
1725 => three warnings about OTP not working
1726 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1728 With the patch, sasl_listmech() also gets called three times. But now
1729 SASL's mech_list option is set to the server_mech specified in the the
1730 authenticator. Or in other words, the answer from sasl_listmech()
1731 gets limited to just the mech you're testing for (which is different
1733 => the return list contains just 'plain' or 'login', 'cram-md5' or
1734 nothing depending on the value of ob->server_mech.
1736 I've just tested the patch: Authentication still works fine,
1737 unavailable mechs specified in the exim configuration are still
1738 caught, and the auth.log warnings about OTP are gone."
1740 PH/25 When debugging is enabled, the contents of the command line are added
1741 to the debugging output, even when log_selector=+arguments is not
1744 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1745 answer is "GNU", and only if the return is "GNU/something" is the answer
1748 PH/27 $acl_verify_message is now set immediately after the failure of a
1749 verification in an ACL, and so is available in subsequent modifiers. In
1750 particular, the message can be preserved by coding like this:
1752 warn !verify = sender
1753 set acl_m0 = $acl_verify_message
1755 Previously, $acl_verify_message was set only while expanding "message"
1756 and "log_message" when a very denied access.
1758 PH/28 Modified OS/os.c-Linux with
1760 -#ifndef OS_LOAD_AVERAGE
1761 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1763 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1764 nomenclature these days.)
1766 PH/29 Installed patch from the Sieve maintainer that adds the options
1767 sieve_useraddress and sieve_subaddress to the redirect router.
1769 PH/30 In these circumstances:
1770 . Two addresses routed to the same list of hosts;
1771 . First host does not offer TLS;
1772 . First host accepts first address;
1773 . First host gives temporary error to second address;
1774 . Second host offers TLS and a TLS session is established;
1775 . Second host accepts second address.
1776 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1777 and peerdn, if requested) that were in fact used only for the second
1780 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1781 attention to any local part prefix or suffix that was matched by the
1782 router that accepted the address. It now behaves in the same way as it
1783 does for delivery: the affixes are removed from the local part unless
1784 rcpt_include_affixes is set on the transport.
1786 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1787 timeout during the DATA phase of an incoming message.
1789 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1790 applied a patch sanctioned by the Sieve maintainer.
1792 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1793 the uid or gid is negative. A case of a negative gid caused this to be
1794 noticed. The fix allows for either to be negative.
1796 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1797 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1800 PH/36 The change PH/12 above was broken. Fixed it.
1802 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1803 the grounds that routing the same address twice would always produce the
1804 same answer. This might have been true once, but it is certainly no
1805 longer true now. Routing a child address may depend on the previous
1806 routing that produced that child. Some complicated redirection strategies
1807 went wrong when messages had multiple recipients, and made Exim's
1808 behaviour dependent on the order in which the addresses were given.
1810 I have moved the duplicate checking until after the routing is complete.
1811 Exim scans the addresses that are assigned to local and remote
1812 transports, and removes any duplicates. This means that more work will be
1813 done, as duplicates will always all be routed, but duplicates are
1814 presumably rare, so I don't expect this is of any significance.
1816 For deliveries to pipes, files, and autoreplies, the duplicate checking
1817 still happens during the routing process, since they are not going to be
1820 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1821 It corrects a timeout issue with spamd. This is Ian's comment: "The
1822 background is that sometimes spamd either never reads data from a
1823 connection it has accepted, or it never writes response data. The exiscan
1824 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1825 blindly assumes that writes won't block so it may never time out."
1827 PH/39 Allow G after quota size as well as K and M.
1829 PH/40 The value set for $authenticated_id in an authenticator may not contain
1830 binary zeroes or newlines because the value is written to log lines and
1831 to spool files. There was no check on this. Now the value is run through
1832 the string_printing() function so that such characters are converted to
1833 printable escape sequences.
1835 PH/41 $message_linecount is a new variable that contains the total number of
1836 lines in the message. Compare $body_linecount, which is the count for the
1839 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1840 bounce and delay warning messages, except in certain special cases, which
1843 (a) An SMTP error message from a remote host;
1844 (b) A message specified in a :fail: redirection;
1845 (c) A message specified in a "fail" command in a system filter;
1846 (d) A message specified in a FAIL return from the queryprogram router;
1847 (e) A message specified by the cannot_route_message router option.
1849 In these cases only, Exim does include the error details in bounce and
1850 warning messages. There are also a few cases where bland messages such
1851 as "unrouteable address" or "local delivery error" are given.
1853 PH/43 $value is now also set for the "else" part of a ${run expansion.
1855 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1856 being worked on, but at least Exim now implements the latest version to
1859 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1860 process to complete was treated as a delivery failure, a timeout while
1861 writing the message to the pipe was logged, but erroneously treated as a
1862 successful delivery. Such timeouts include transport filter timeouts. For
1863 consistency with the overall process timeout, these timeouts are now
1864 treated as errors, giving rise to delivery failures by default. However,
1865 there is now a new Boolean option for the pipe transport called
1866 timeout_defer, which, if set TRUE, converts the failures into defers for
1867 both kinds of timeout. A transport filter timeout is now identified in
1870 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1871 systems where "make" and "gmake" are different, calling "gmake" at top
1872 level broke things. I've arranged for the value of $(MAKE) to be passed
1873 from the Makefile to this script so that it can call the same version of
1877 A note about Exim versions 4.44 and 4.50
1878 ----------------------------------------
1880 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1881 changes of various kinds. As a consequence, a big documentation update was
1882 needed. This delayed the release for rather longer than seemed good, especially
1883 in the light of a couple of (minor) security issues. Therefore, the changes
1884 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1885 release. So 4.44 and 4.50 are in effect two different branches that both start
1888 I have left the 4.50 change log unchanged; it contains all the changes since
1889 4.43. The change log for 4.44 is below; many of its items are identical to
1890 those for 4.50. This seems to be the most sensible way to preserve the
1891 historical information.
1897 1. Minor wording change to the doc/README.SIEVE file.
1899 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1900 computation of the current number of files was incorrect.
1902 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1903 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1906 4. Give more explanation in the error message when the command for a transport
1907 filter fails to execute.
1909 5. There are several places where Exim runs a non-Exim command in a
1910 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1911 was being done only for the command run by the queryprogram router. It is
1912 now done for all such subprocesses. The other cases are: ${run, transport
1913 filters, and the commands run by the lmtp and pipe transports.
1915 6. Added CONFIGURE_GROUP build-time option.
1917 7. Some older OS have a limit of 256 on the maximum number of file
1918 descriptors. Exim was using setrlimit() to set 1000 as a large value
1919 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1920 systems. I've change it so that if it can't get 1000, it tries for 256.
1922 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1923 was an oversight, and furthermore, ever since the addition of extra
1924 controls (e.g. 4.43/32), the checks on when to allow different forms of
1925 "control" were broken. There should now be diagnostics for all cases when a
1926 control that does not make sense is encountered.
1928 9. Added the /retain_sender option to "control=submission".
1930 10. $recipients is now available in the predata ACL (oversight).
1932 11. Tidy the search cache before the fork to do a delivery from a message
1933 received from the command line. Otherwise the child will trigger a lookup
1934 failure and thereby defer the delivery if it tries to use (for example) a
1935 cached ldap connection that the parent has called unbind on.
1937 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1938 of $address_data from the recipient verification was clobbered by the
1939 sender verification.
1941 13. The value of address_data from a sender verification is now available in
1942 $sender_address_data in subsequent conditions in the ACL statement.
1944 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1946 15. Added a new option "connect=<time>" to callout options, to set a different
1949 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1950 was its contents. (It was OK if the option was not defined at all.)
1952 17. A "Completed" log line is now written for messages that are removed from
1953 the spool by the -Mrm option.
1955 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1956 information about exactly what failed.
1958 19. Added -dd to debug only the daemon process.
1960 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1961 handles timeouts, both on the server side and network timeouts. Renamed the
1962 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1964 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1965 It is now set to "smtps".
1967 22. $host_address is now set to the target address during the checking of
1968 ignore_target_hosts.
1970 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1971 being passed; this would have caused $sender_host_name to have been used if
1972 matching the list had actually called for a host name (not very likely,
1973 since this list is usually IP addresses). A host name is now passed as
1976 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1977 code that specifies a non-restarting handler (typically sigaction() in
1978 modern systems) in an attempt to fix a rare and obscure crash bug.
1980 25. Narrowed the window for a race in the daemon that could cause it to ignore
1981 SIGCHLD signals. This is not a major problem, because they are used only to
1982 wake it up if nothing else does.
1984 26. A malformed maildirsize file could cause Exim to calculate negative values
1985 for the mailbox size or file count. Odd effects could occur as a result.
1986 The maildirsize information is now recalculated if the size or filecount
1989 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1990 support for a long time. Removed HAVE_SYS_VFS_H.
1992 28. Installed the latest version of exipick from John Jetmore.
1994 29. In an address list, if the pattern was not a regular expression, an empty
1995 subject address (from a bounce message) matched only if the pattern was an
1996 empty string. Non-empty patterns were not even tested. This was the wrong
1997 because it is perfectly reasonable to use an empty address as part of a
1998 database query. An empty address is now tested by patterns that are
1999 lookups. However, all the other forms of pattern expect the subject to
2000 contain a local part and a domain, and therefore, for them, an empty
2001 address still always fails if the pattern is not itself empty.
2003 30. Exim went into a mad DNS loop when attempting to do a callout where the
2004 host was specified on an smtp transport, and looking it up yielded more
2005 than one IP address.
2007 31. Re-factored the code for checking spool and log partition space into a
2008 function that finds that data and another that does the check. The former
2009 is then used to implement four new variables: $spool_space, $log_space,
2010 $spool_inodes, and $log_inodes.
2012 32. The RFC2047 encoding function was originally intended for short strings
2013 such as real names; it was not keeping to the 75-character limit for
2014 encoded words that the RFC imposes. It now respects the limit, and
2015 generates multiple encoded words if necessary. To be on the safe side, I
2016 have increased the buffer size for the ${rfc2047: expansion operator from
2019 33. It is now permitted to omit both strings after an "if" condition; if the
2020 condition is true, the result is "true". As before, when the second string
2021 is omitted, a false condition yields an empty string. This makes it less
2022 cumbersome to write custom ACL and router conditions.
2024 34. Failure to deliver a bounce message always caused it to be frozen, even if
2025 there was an errors_to setting on the router. The errors_to setting is now
2028 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2029 canonical form (fully expanded) before being placed in
2030 $sender_host_address.
2032 36. The table in the code that translates DNS record types into text (T_A to
2033 "A" for instance) was missing entries for NS and CNAME. It is just possible
2034 that this could have caused confusion if both these types were looked up
2035 for the same domain, because the text type is used as part of Exim's
2036 per-process caching. But the chance of anyone hitting this buglet seems
2039 37. The dnsdb lookup has been extended in a number of ways.
2041 (1) There is a new type, "zns", which walks up the domain tree until it
2042 finds some nameserver records. It should be used with care.
2044 (2) There is a new type, "mxh", which is like "mx" except that it returns
2045 just the host names, not the priorities.
2047 (3) It is now possible to give a list of domains (or IP addresses) to be
2048 looked up. The behaviour when one of the lookups defers can be
2049 controlled by a keyword.
2051 (4) It is now possible to specify the separator character for use when
2052 multiple records are returned.
2054 38. The dnslists ACL condition has been extended: it is now possible to supply
2055 a list of IP addresses and/or domains to be looked up in a particular DNS
2058 39. Added log_selector=+queue_time_overall.
2060 40. When running the queue in the test harness, wait just a tad after forking a
2061 delivery process, to get repeatability of debugging output.
2063 41. Include certificate and key file names in error message when GnuTLS fails
2064 to set them up, because the GnuTLS error message doesn't include the name
2065 of the failing file when there is a problem reading it.
2067 42. Allow both -bf and -bF in the same test run.
2069 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2071 44. The "Exiscan patch" is now merged into the mainline Exim source.
2073 45. Sometimes the final signoff response after QUIT could fail to get
2074 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2075 < 0 before doing a fflush(). This bug looks as though it goes back to the
2076 introduction of TLS in release 3.20, but "sometimes" must have been rare
2077 because the tests only now provoked it.
2079 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2080 (this can affect the format of dates).
2082 47. exim_tidydb, when checking for the continued existence of a message for
2083 which it has found a message-specific retry record, was not finding
2084 messages that were in split spool directories. Consequently, it was
2085 deleting retry records that should have stayed in existence.
2087 48. Steve fixed some bugs in eximstats.
2089 49. The SPA authentication driver was not abandoning authentication and moving
2090 on to the next authenticator when an expansion was forced to fail,
2091 contradicting the general specification for all authenticators. Instead it
2092 was generating a temporary error. It now behaves as specified.
2094 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2095 (the order specifies the preference for clients). The order is now AES256,
2096 AES128, 3DES, ARCFOUR128.
2098 51. Small patch to Sieve code - explicitly set From: when generating an
2101 52. Exim crashed if a remote delivery caused a very long error message to be
2102 recorded - for instance if somebody sent an entire SpamAssassin report back
2103 as a large number of 550 error lines. This bug was coincidentally fixed by
2104 increasing the size of one of Exim's internal buffers (big_buffer) that
2105 happened as part of the Exiscan merge. However, to be on the safe side, I
2106 have made the code more robust (and fixed the comments that describe what
2109 53. Now that there can be additional text after "Completed" in log lines (if
2110 the queue_time_overall log selector is set), a one-byte patch to exigrep
2111 was needed to allow it to recognize "Completed" as not the last thing in
2114 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2115 patch that reportedly fixes this has been added. I am not expert enough to
2116 create a test for it. This is what the patch creator wrote:
2118 "I found a little strange behaviour of ldap code when working with
2119 Windows 2003 AD Domain, where users was placed in more than one
2120 Organization Units. When I tried to give exim partial DN, the exit code
2121 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2122 But simultaneously result of request was absolutely normal ldap result,
2123 so I produce this patch..."
2125 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2126 REFERENCE, so I have modified the code to exclude the patch when that macro
2129 55. Some experimental protocols are using DNS PTR records for new purposes. The
2130 keys for these records are domain names, not reversed IP addresses. The
2131 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2132 leaves it alone. Component reversal etc. now happens only for IP addresses.
2135 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2137 57. Double the size of the debug message buffer (to 2048) so that more of very
2138 long debug lines gets shown.
2140 58. The exicyclog utility now does better if the number of log files to keep
2141 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2143 59. Two changes related to the smtp_active_hostname option:
2145 (1) $smtp_active_hostname is now available as a variable.
2146 (2) The default for smtp_banner uses $smtp_active_hostname instead
2147 of $primary_hostname.
2149 60. The host_aton() function is supposed to be passed a string that is known
2150 to be a valid IP address. However, in the case of IPv6 addresses, it was
2151 not checking this. This is a hostage to fortune. Exim now panics and dies
2152 if the condition is not met. A case was found where this could be provoked
2153 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2154 components; fortuitously, this particular loophole had already been fixed
2155 by change 4.50/55 above.
2157 If there are any other similar loopholes, the new check in host_aton()
2158 itself should stop them being exploited. The report I received stated that
2159 data on the command line could provoke the exploit when Exim was running as
2160 exim, but did not say which command line option was involved. All I could
2161 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2162 running as the user.
2165 61. There was a buffer overflow vulnerability in the SPA authentication code
2166 (which came originally from the Samba project). I have added a test to the
2167 spa_base64_to_bits() function which I hope fixes it.
2170 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2171 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2173 63. The daemon start-up calls getloadavg() while still root for those OS that
2174 need the first call to be done as root, but it missed one case: when
2175 deliver_queue_load_max is set with deliver_drop_privilege. This is
2176 necessary for the benefit of the queue runner, because there is no re-exec
2177 when deliver_drop_privilege is set.
2179 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2180 This has been fixed.
2182 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2183 was in use, was not putting the data itself into the right store pool;
2184 consequently, it could be overwritten for a subsequent message in the same
2185 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2188 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2190 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2191 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2192 the function were treating the return as a boolean value, which happened to
2193 work because 0=false and not-0=true, but is not correct code.
2195 68. The host_aton() function was not handling scoped IPv6 addresses (those
2196 with, for example, "%eth0" on the end) correctly.
2198 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2199 negated items (that is, ~something) in unsigned ints. Some compilers
2200 apparently mutter when there is no cast.
2202 70. If an address verification called from an ACL failed, and did not produce a
2203 user-specific message (i.e. there was only a "system" message), nothing was
2204 put in $acl_verify_message. In this situation, it now puts the system
2207 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2208 session; change 4.31/43 added the unwanted input to the log line - except
2209 that it did not do this in the start of session case. It now does.
2211 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2212 This is wrong and can cause the other end to generate a synchronization
2213 error if it is another Exim or anything else that does the synchronization
2214 check. A QUIT command is no longer sent after a timeout.
2216 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2217 during host lookups.
2219 74. The defer_ok option of callout verification was not working if it was used
2220 when verifying addresses in header lines, that is, for this case:
2222 verify = header_sender/callout=defer_ok
2224 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2225 those file descriptors could be used for SMTP connections. If anything
2226 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2227 could be sent to the SMTP client, causing chaos. The daemon now opens
2228 stdin, stdout, and stderr to /dev/null when it puts itself into the
2231 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2232 log by default. The user can override this with suitable Perl magic.
2234 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2235 the log message when discard triggers, was not working for the DATA ACL or
2236 for the non-SMTP ACL.
2238 78. Error message wording change in sieve.c.
2240 79. If smtp_accept_max_per_host was set, the number of connections could be
2241 restricted to fewer than expected, because the daemon was trying to set up
2242 a new connection before checking whether the processes handling previous
2243 connections had finished. The check for completed processes is now done
2244 earlier. On busy systems, this bug wouldn't be noticed because something
2245 else would have woken the daemon, and it would have reaped the completed
2248 80. If a message was submitted locally by a user whose login name contained one
2249 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2250 It caused a spool format error. I have fixed the spool reading code. A
2251 related problem was that the "from" clause in the Received: line became
2252 illegal because of the space(s). It is now covered by ${quote_local_part.
2254 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2257 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2259 83. Patch from Sieve maintainer to fix unterminated string problem in
2260 "vacation" handling.
2262 84. Some minor changes to the Linux configuration files to help with other
2263 OS variants using glibc.
2265 85. One more patch for Sieve to update vacation handling to latest spec.
2268 ----------------------------------------------------
2269 See the note above about the 4.44 and 4.50 releases.
2270 ----------------------------------------------------
2276 1. Change 4.43/35 introduced a bug that caused file counts to be
2277 incorrectly computed when quota_filecount was set in an appendfile
2280 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2281 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2284 3. Give more explanation in the error message when the command for a transport
2285 filter fails to execute.
2287 4. There are several places where Exim runs a non-Exim command in a
2288 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2289 was being done only for the command run by the queryprogram router. It is
2290 now done for all such subprocesses. The other cases are: ${run, transport
2291 filters, and the commands run by the lmtp and pipe transports.
2293 5. Some older OS have a limit of 256 on the maximum number of file
2294 descriptors. Exim was using setrlimit() to set 1000 as a large value
2295 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2296 systems. I've change it so that if it can't get 1000, it tries for 256.
2298 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2299 was an oversight, and furthermore, ever since the addition of extra
2300 controls (e.g. 4.43/32), the checks on when to allow different forms of
2301 "control" were broken. There should now be diagnostics for all cases when a
2302 control that does not make sense is encountered.
2304 7. $recipients is now available in the predata ACL (oversight).
2306 8. Tidy the search cache before the fork to do a delivery from a message
2307 received from the command line. Otherwise the child will trigger a lookup
2308 failure and thereby defer the delivery if it tries to use (for example) a
2309 cached ldap connection that the parent has called unbind on.
2311 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2312 of $address_data from the recipient verification was clobbered by the
2313 sender verification.
2315 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2316 was its contents. (It was OK if the option was not defined at all.)
2318 11. A "Completed" log line is now written for messages that are removed from
2319 the spool by the -Mrm option.
2321 12. $host_address is now set to the target address during the checking of
2322 ignore_target_hosts.
2324 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2325 being passed; this would have caused $sender_host_name to have been used if
2326 matching the list had actually called for a host name (not very likely,
2327 since this list is usually IP addresses). A host name is now passed as
2330 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2331 code that specifies a non-restarting handler (typically sigaction() in
2332 modern systems) in an attempt to fix a rare and obscure crash bug.
2334 15. Narrowed the window for a race in the daemon that could cause it to ignore
2335 SIGCHLD signals. This is not a major problem, because they are used only to
2336 wake it up if nothing else does.
2338 16. A malformed maildirsize file could cause Exim to calculate negative values
2339 for the mailbox size or file count. Odd effects could occur as a result.
2340 The maildirsize information is now recalculated if the size or filecount
2343 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2344 support for a long time. Removed HAVE_SYS_VFS_H.
2346 18. Updated exipick to current release from John Jetmore.
2348 19. Allow an empty sender to be matched against a lookup in an address list.
2349 Previously the only cases considered were a regular expression, or an
2352 20. Exim went into a mad DNS lookup loop when doing a callout where the
2353 host was specified on the transport, if the DNS lookup yielded more than
2356 21. The RFC2047 encoding function was originally intended for short strings
2357 such as real names; it was not keeping to the 75-character limit for
2358 encoded words that the RFC imposes. It now respects the limit, and
2359 generates multiple encoded words if necessary. To be on the safe side, I
2360 have increased the buffer size for the ${rfc2047: expansion operator from
2363 22. Failure to deliver a bounce message always caused it to be frozen, even if
2364 there was an errors_to setting on the router. The errors_to setting is now
2367 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2368 canonical form (fully expanded) before being placed in
2369 $sender_host_address.
2371 24. Updated eximstats to version 1.33
2373 25. Include certificate and key file names in error message when GnuTLS fails
2374 to set them up, because the GnuTLS error message doesn't include the name
2375 of the failing file when there is a problem reading it.
2377 26. Expand error message when OpenSSL has problems setting up cert/key files.
2380 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2381 (this can affect the format of dates).
2383 28. exim_tidydb, when checking for the continued existence of a message for
2384 which it has found a message-specific retry record, was not finding
2385 messages that were in split spool directories. Consequently, it was
2386 deleting retry records that should have stayed in existence.
2388 29. eximstats updated to version 1.35
2389 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2390 1.35 - bugfix such that pie charts by volume are generated correctly
2392 30. The SPA authentication driver was not abandoning authentication and moving
2393 on to the next authenticator when an expansion was forced to fail,
2394 contradicting the general specification for all authenticators. Instead it
2395 was generating a temporary error. It now behaves as specified.
2397 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2398 (the order specifies the preference for clients). The order is now AES256,
2399 AES128, 3DES, ARCFOUR128.
2401 31. Small patch to Sieve code - explicitly set From: when generating an
2404 32. Exim crashed if a remote delivery caused a very long error message to be
2405 recorded - for instance if somebody sent an entire SpamAssassin report back
2406 as a large number of 550 error lines. This bug was coincidentally fixed by
2407 increasing the size of one of Exim's internal buffers (big_buffer) that
2408 happened as part of the Exiscan merge. However, to be on the safe side, I
2409 have made the code more robust (and fixed the comments that describe what
2412 33. Some experimental protocols are using DNS PTR records for new purposes. The
2413 keys for these records are domain names, not reversed IP addresses. The
2414 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2415 leaves it alone. Component reversal etc. now happens only for IP addresses.
2418 34. The host_aton() function is supposed to be passed a string that is known
2419 to be a valid IP address. However, in the case of IPv6 addresses, it was
2420 not checking this. This is a hostage to fortune. Exim now panics and dies
2421 if the condition is not met. A case was found where this could be provoked
2422 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2423 components; fortuitously, this particular loophole had already been fixed
2424 by change 4.50/55 or 4.44/33 above.
2426 If there are any other similar loopholes, the new check in host_aton()
2427 itself should stop them being exploited. The report I received stated that
2428 data on the command line could provoke the exploit when Exim was running as
2429 exim, but did not say which command line option was involved. All I could
2430 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2431 running as the user.
2434 35. There was a buffer overflow vulnerability in the SPA authentication code
2435 (which came originally from the Samba project). I have added a test to the
2436 spa_base64_to_bits() function which I hope fixes it.
2439 36. The daemon start-up calls getloadavg() while still root for those OS that
2440 need the first call to be done as root, but it missed one case: when
2441 deliver_queue_load_max is set with deliver_drop_privilege. This is
2442 necessary for the benefit of the queue runner, because there is no re-exec
2443 when deliver_drop_privilege is set.
2445 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2446 was in use, was not putting the data itself into the right store pool;
2447 consequently, it could be overwritten for a subsequent message in the same
2448 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2451 38. Sometimes the final signoff response after QUIT could fail to get
2452 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2453 < 0 before doing a fflush(). This bug looks as though it goes back to the
2454 introduction of TLS in release 3.20, but "sometimes" must have been rare
2455 because the tests only now provoked it.
2461 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2462 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2463 Now it returns an integer. A number of calls were still expecting a T/F
2464 return. Fortuitously, in all cases, the tests worked in OK situations,
2465 which is the norm. However, things would have gone wrong on any write
2466 failures on the smtp file descriptor. This function is used when sending
2467 messages over SMTP and also when doing verify callouts.
2469 2. When Exim is called to do synchronous delivery of a locally submitted
2470 message (the -odf or -odi options), it no longer closes stderr before doing
2473 3. Implemented the mua_wrapper option.
2475 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2477 5. Implemented the functions header_remove(), header_testname(),
2478 header_add_at_position(), and receive_remove_recipient(), and exported them
2481 6. If an ACL "warn" statement specified the addition of headers, Exim already
2482 inserted X-ACL-Warn: at the start if there was no header name. However, it
2483 was not making this test for the second and subsequent header lines if
2484 there were newlines in the string. This meant that an invalid header could
2485 be inserted if Exim was badly configured.
2487 7. Allow an ACL "warn" statement to add header lines at the start or after all
2488 the Received: headers, as well as at the end.
2490 8. Added the rcpt_4xx retry error code.
2492 9. Added postmaster_mailfrom=xxx to callout verification option.
2494 10. Added mailfrom=xxxx to the callout verification option, for verify=
2497 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2498 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2500 12. Inserted some casts to stop certain compilers warning when using pointer
2501 differences as field lengths or precisions in printf-type calls (mostly
2502 affecting debugging statements).
2504 13. Added optional readline() support for -be (dynamically loaded).
2506 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2507 same clock tick as a message's arrival, so that its received time was the
2508 same as the "first fail" time on the retry record, and that message
2509 remained on the queue past the ultimate address timeout, every queue runner
2510 would try a delivery (because it was past the ultimate address timeout) but
2511 after another failure, the ultimate address timeout, which should have then
2512 bounced the address, did not kick in. This was a "< instead of <=" error;
2513 in most cases the first failure would have been in the next clock tick
2514 after the received time, and all would be well.
2516 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2517 being recognized when the domain list was tested by the match_domain
2518 condition in an expansion string.
2520 16. Added the ${str2b64: operator.
2522 17. Exim was always calling setrlimit() to set a large limit for the number of
2523 processes, without checking whether the existing limit was already
2524 adequate. (It did check for the limit on file descriptors.) Furthermore,
2525 errors from getrlimit() and setrlimit() were being ignored. Now they are
2526 logged to the main and panic logs, but Exim does carry on, to try to do its
2527 job under whatever limits there are.
2529 18. Imported PCRE 5.0.
2531 19. Trivial typo in log message " temporarily refused connection" (the leading
2534 20. If the log selector return_path_on_delivery was set and an address was
2535 redirected to /dev/null, the delivery process crashed because it assumed
2536 that a return path would always be set for a "successful" delivery. In this
2537 case, the whole delivery is bypassed as an optimization, and therefore no
2540 21. Internal re-arrangement: the function for sending a challenge and reading
2541 a response while authentication was assuming a zero-terminated challenge
2542 string. It's now changed to take a pointer and a length, to allow for
2543 binary data in such strings.
2545 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2547 23. Exim was not respecting finduser_retries when seeking the login of the
2548 uid under which it was called; it was always trying 10 times. (The default
2549 setting of finduser_retries is zero.) Also, it was sleeping after the final
2550 failure, which is pointless.
2552 24. Implemented tls_on_connect_ports.
2554 25. Implemented acl_smtp_predata.
2556 26. If the domain in control=submission is set empty, Exim assumes that the
2557 authenticated id is a complete email address when it generates From: or
2558 Sender: header lines.
2560 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2561 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2562 chown and chgrp in /bin and hostname in /usr/bin.
2564 28. Exim was keeping the "process log" file open after each use, just as it
2565 does for the main log. This opens the possibility of it remaining open for
2566 long periods when the USR1 signal hits a daemon. Occasional processlog
2567 errors were reported, that could have been caused by this. Anyway, it seems
2568 much more sensible not to leave this file open at all, so that is what now
2571 29. The long-running daemon process does not normally write to the log once it
2572 has entered its main loop, and it closes the log before doing so. This is
2573 so that log files can straightforwardly be renamed and moved. However,
2574 there are a couple of unusual error situations where the daemon does write
2575 log entries, and I had neglected to close the log afterwards.
2577 30. The text of an SMTP error response that was received during a remote
2578 delivery was being truncated at 512 bytes. This is too short for some of
2579 the long messages that one sometimes sees. I've increased the limit to
2582 31. It is now possible to make retry rules that apply only when a message has a
2583 specific sender, in particular, an empty sender.
2585 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2586 it possible to be selective about when SMTP synchronization is enforced.
2588 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2590 32. Implemented hosts_connection_nolog.
2592 33. Added an ACL for QUIT.
2594 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2597 35. Added mailbox_size and mailbox_filecount to appendfile.
2599 36. Added control = no_multiline_responses to ACLs.
2601 37. There was a bug in the logic of the code that waits for the clock to tick
2602 in the case where the clock went backwards by a substantial amount such
2603 that the microsecond fraction of "now" was more than the microsecond
2604 fraction of "then" (but the whole seconds number was less).
2606 38. Added support for the libradius Radius client library this is found on
2607 FreeBSD (previously only the radiusclient library was supported).
2613 1. When certain lookups returned multiple values in the form name=value, the
2614 quoting of the values was not always being done properly. Specifically:
2615 (a) If the value started with a double quote, but contained no whitespace,
2617 (b) If the value contained whitespace other than a space character (i.e.
2618 tabs or newlines or carriage returns) it was not quoted.
2619 This fix has been applied to the mysql and pgsql lookups by writing a
2620 separate quoting function and calling it from the lookup code. The fix
2621 should probably also be applied to nisplus, ibase and oracle lookups, but
2622 since I cannot test any of those, I have not disturbed their existing code.
2624 2. A hit in the callout cache for a specific address caused a log line with no
2625 reason for rejecting RCPT. Now it says "Previous (cached) callout
2626 verification failure".
2628 3. There was an off-by-one bug in the queryprogram router. An over-long
2629 return line was truncated at 256 instead of 255 characters, thereby
2630 overflowing its buffer with the terminating zero. As well as fixing this, I
2631 have increased the buffer size to 1024 (and made a note to document this).
2633 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2634 when Exim is waiting for an SMTP response from a remote server, Exim
2635 restarts its select() call on the socket, thereby resetting its timeout.
2636 This is not a problem when such interrupts are rare. Somebody set up a cron
2637 job to run exiwhat every 2 minutes, which is less than the normal select()
2638 timeout (5 or 10 minutes). This meant that the select() timeout never
2639 kicked in because it was always reset. I have fixed this by comparing the
2640 time when an interrupt arrives with the time at the start of the first call
2641 to select(). If more time than the timeout has elapsed, the interrupt is
2642 treated as a timeout.
2644 5. Some internal re-factoring in preparation for the addition of Sieve
2645 extensions (by MH). In particular, the "personal" test is moved to a
2646 separate function, and given an option for scanning Cc: and Bcc: (which is
2647 not set for Exim filters).
2649 6. When Exim created an email address using the login of the caller as the
2650 local part (e.g. when creating a From: or Sender: header line), it was not
2651 quoting the local part when it contained special characters such as @.
2653 7. Installed new OpenBSD configuration files.
2655 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2656 try to make them clearer.
2658 9. Callout options, other than the timeout value, were being ignored when
2659 verifying sender addresses in header lines. For example, when using
2661 verify = header_sender/callout=no_cache
2663 the cache was (incorrectly) being used.
2665 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2666 only the Cygwin environment.
2668 11. When return_path_on_delivery was set as a log selector, if different remote
2669 addresses in the same message used different return paths and parallel
2670 remote delivery occurred, the wrong values would sometimes be logged.
2671 (Whenever a remote delivery process finished, the return path value from
2672 the most recently started remote delivery process was logged.)
2674 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2675 lines when AUTH and/or TLS are in use. This is the "received protocol"
2676 field. Exim used to use "asmtp" for authenticated SMTP, without any
2677 indication (in the protocol name) for TLS use. Now it follows the RFC and
2678 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2679 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2680 names appear in log lines as well as in Received: header lines.
2682 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2683 extensions, and comparison tests, and to fix some bugs.
2685 14. Changes to the "personal" filter test:
2687 (1) The test was buggy in that it was just doing the equivalent of
2688 "contains" tests on header lines. For example, if a user's address was
2689 anne@some.where, the "personal" test would incorrectly be true for
2691 To: susanne@some.where
2693 This test is now done by extracting each address from the header in turn,
2694 and checking the entire address. Other tests that are part of "personal"
2695 are now done using regular expressions (for example, to check local parts
2696 of addresses in From: header lines).
2698 (2) The list of non-personal local parts in From: addresses has been
2699 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2700 taken from the Sieve specification recommendations.
2702 (3) If the message contains any header line starting with "List-" it is
2703 treated as non-personal.
2705 (4) The test for "circular" in the Subject: header line has been removed
2706 because it now seems ill-conceived.
2708 15. Minor typos in src/EDITME comments corrected.
2710 16. Installed latest exipick from John Jetmore.
2712 17. If headers_add on a router specified a text string that was too long for
2713 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2714 of string_sprintf() is now avoided.
2716 18. $message_body_size was not set (it was always zero) when running the DATA
2717 ACL and the local_scan() function.
2719 19. For the "mail" command in an Exim filter, no default was being set for
2720 the once_repeat time, causing a random time value to be used if "once" was
2721 specified. (If the value happened to be <= 0, no repeat happened.) The
2722 default is now 0s, meaning "never repeat". The "vacation" command was OK
2723 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2724 (I found it when inspecting the code).
2726 20. There is now an overall timeout for performing a callout verification. It
2727 defaults to 4 times the callout timeout, which applies to individual SMTP
2728 commands during the callout. The overall timeout applies when there is more
2729 than one host that can be tried. The timeout is checked before trying the
2730 next host. This prevents very long delays if there are a large number of
2731 hosts and all are timing out (e.g. when the network connections are timing
2732 out). The value of the overall timeout can be changed by specifying an
2733 additional sub-option for "callout", called "maxwait". For example:
2735 verify = sender/callout=5s,maxwait=20s
2737 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2738 to the end before writing, but this should make it even safer).
2740 22. Exim was forgetting that it had advertised PIPELINING for the second and
2741 subsequent messages on an SMTP connection. It was also not resetting its
2742 memory on STARTTLS and an internal HELO.
2744 23. When Exim logs an SMTP synchronization error within a session, it now
2745 records whether PIPELINING has been advertised or not.
2747 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2748 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2749 rather than long int.
2751 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2753 26. Added the never_mail option to autoreply.
2759 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2760 crash if the getsockname() call failed; this can happen if a connection is
2761 closed very soon after it is established. The problem was simply in the
2762 order in which certain operations were done, causing Exim to try to write
2763 to the SMTP stream before it had set up the file descriptor. The bug has
2764 been fixed by making things happen in the correct order.
2770 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2771 before the connection was closed, thus losing the rejection response.
2773 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2774 some early Solaris releases, but causes trouble in current releases where
2775 socklen_t is defined.
2777 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2780 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2783 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2784 that releasing the top of it at the end releases what was used for sub-
2785 expansions (unless the block got too big). However, discard this block if
2786 the first thing is a variable or header, so that we can use its block when
2787 it is dynamic (useful for very large $message_headers, for example).
2789 6. Lookups now cache *every* query, not just the most recent. A new, separate
2790 store pool is used for this. It can be recovered when all lookup caches are
2791 flushed. Lookups now release memory at the end of their result strings.
2792 This has involved some general refactoring of the lookup sources.
2794 7. Some code has been added to the store_xxx() functions to reduce the amount
2795 of flapping under certain conditions.
2797 8. log_incoming_interface used to affect only the <= reception log lines. Now
2798 it causes the local interface and port to be added to several more SMTP log
2799 lines, for example "SMTP connection from", and rejection lines.
2801 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2803 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2805 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2806 could be overwritten at the end of the current message (or the start of a
2807 new message if it was set in a HELO ACL). The value is now preserved for
2808 the duration of the SMTP connection.
2810 12. If a transport had a headers_rewrite setting, and a matching header line
2811 contained an unqualified address, that address was qualified, even if it
2812 did not match any rewriting rules. The underlying bug was that the values
2813 of the flags that permit the existence of unqualified sender and recipient
2814 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2815 non-local messages, and by -bnq for local messages) were not being
2816 preserved with the message after it was received.
2818 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2819 "next input=" as part of the text comprising the host identity instead of
2820 the correct text. The code was using the same buffer for two different
2821 strings. However, depending on which order the printing function evaluated
2822 its arguments, the bug did not always show up. Under Linux, for example, my
2823 test suite worked just fine.
2825 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2826 doesn't work with some older versions of Perl. It has been changed to "my",
2827 which in any case is probably the better facility to use.
2829 15. A really picky compiler found some instances of statements for creating
2830 error messages that either had too many or two few arguments for the format
2833 16. The size of the buffer for calls to the DNS resolver has been increased
2834 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2835 for addresses that have a lot of PTR records. This alleviates a problem; it
2836 does not fully solve it.
2838 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2839 buffer now truncates the list and logs the incident, which is the same
2840 action as happens when Exim is looking up a host name and its aliases.
2841 Previously in this situation something unpredictable would happen;
2842 sometimes it was "internal error: store_reset failed".
2844 18. If a server dropped the connection unexpectedly when an Exim client was
2845 using GnuTLS and trying to read a response, the client delivery process
2846 crashed while trying to generate an error log message.
2848 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2849 string, for example:
2851 warn message = H1: something\nH2: something
2853 the text was added as a single header line from Exim's point of view
2854 though it ended up OK in the delivered message. However, searching for the
2855 second and subsequent header lines using $h_h2: did not work. This has been
2856 fixed. Similarly, if a system filter added multiple headers in this way,
2857 the routers could not see them.
2859 20. Expanded the error message when iplsearch is called with an invalid key to
2860 suggest using net-iplsearch in a host list.
2862 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2863 ACLs are no longer actually imposed (and a message to that effect is
2866 22. If a "gecos" field in a passwd entry contained escaped characters, in
2867 particular, if it contained a \" sequence, Exim got it wrong when building
2868 a From: or a Sender: header from that name. A second bug also caused
2869 incorrect handling when an unquoted " was present following a character
2870 that needed quoting.
2872 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2873 was not being matched caselessly.
2875 24. Arranged for all hyphens in the exim.8 source to be escaped with
2878 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2879 properly. Recipient callout cache records were still being keyed to include
2880 the sender, even when use_sender was set false. This led to far more
2881 callouts that were necessary. The sender is no longer included in the key
2882 when use_sender is false.
2884 26. Added "control = submission" modifier to ACLs.
2886 27. Added the ${base62d: operator to decode base 62 numbers.
2888 28. dnsdb lookups can now access SRV records.
2890 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2891 the configuration file.
2893 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2894 (-v) mode. This makes the output for a verbose queue run more intelligible.
2896 31. Added a use_postmaster feature to recipient callouts.
2898 32. Added the $body_zerocount variable, containing the number of binary zero
2899 bytes in the message body.
2901 33. The time of last modification of the "new" subdirectory is now used as the
2902 "mailbox time last read" when there is a quota error for a maildir
2905 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2907 35. Added +ignore_unknown as a special item in host lists.
2909 36. Code for decoding IPv6 addresses in host lists is now included, even if
2910 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2911 address was recognized as an IP address, but was then not correctly decoded
2912 into binary, causing unexpected and incorrect effects when compared with
2919 1. Very minor rewording of debugging text in manualroute to say "list of
2920 hosts" instead of "hostlist".
2922 2. If verify=header_syntax was set, and a header line with an unqualified
2923 address (no domain) and a large number of spaces between the end of the
2924 name and the colon was received, the reception process suffered a buffer
2925 overflow, and (when I tested it) crashed. This was caused by some obsolete
2926 code that should have been removed. The fix is to remove it!
2928 3. When running in the test harness, delay a bit after writing a bounce
2929 message to get a bit more predictability in the log output.
2931 4. Added a call to search_tidyup() just before forking a reception process. In
2932 theory, someone could use a lookup in the expansion of smtp_accept_max_
2933 per_host which, without the tidyup, could leave open a database connection.
2935 5. Added the variables $recipient_data and $sender_data which get set from a
2936 lookup success in an ACL "recipients" or "senders" condition, or a router
2937 "senders" option, similar to $domain_data and $local_part_data.
2939 6. Moved the writing of debug_print from before to after the "senders" test
2942 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2943 problems for message scanning, either using a data ACL, or using
2944 local_scan() because the Received: header was not generated till after they
2945 were called (in order to set the time as the time of reception completion).
2946 I have revised the way this works. The header is now generated after the
2947 body is received, but before the ACL or local_scan() are called. After they
2948 are run, the timestamp in the header is updated.
2954 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2955 before starting a queue runner without re-exec. This happened only when
2956 deliver_drop_privilege was set or when the Exim user was set to root. The
2957 effect of the bug was that timeouts during subsequent deliveries caused
2958 crashes instead of being properly handled. The handler is now left at its
2959 default (and expected) setting.
2961 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2962 message, again when deliver_drop_privilege is set or Exim is run as root.
2963 The bug described in (1) was not present in this case, but the tidying up
2964 of the other signals was missing. I have made the two cases consistent.
2966 3. The ignore_target_hosts setting on a manualroute router was being ignored
2967 for hosts that were looked up using the /MX notation.
2969 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2972 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2973 operated on the sender address. After changing the $sender_address to <>
2974 for the sender address verify, Exim was re-instated it as the original
2975 (before rewriting) address, but remembering that it had rewritten it, so it
2976 wasn't rewriting it again. This bug also had the effect of breaking the
2977 sender address verification caching when the sender address was rewritten.
2979 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2980 This has been changed so that if the ip literal address matches
2981 ignore_target_hosts, the router declines.
2983 7. Added expansion conditions match_domain, match_address, and match_local_
2984 part (NOT match_host).
2986 8. The placeholder for the Received: header didn't have a length field set.
2988 9. Added code to Exim itself and to exim_lock to test for a specific race
2989 condition that could lead to file corruption when using MBX delivery. The
2990 issue is with the lockfile that is created in /tmp. If this file is removed
2991 after a process has opened it but before that process has acquired a lock,
2992 there is the potential for a second process to recreate the file and also
2993 acquire a lock. This could lead to two Exim processes writing to the file
2994 at the same time. The added code performs the same test as UW imapd; it
2995 checks after acquiring the lock that its file descriptor still refers to
2996 the same named file.
2998 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2999 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3002 11. Added the smtp_active_hostname option. If used, this will typically be made
3003 to depend on the incoming interface address. Because $interface_address is
3004 not set up until the daemon has forked a reception process, error responses
3005 that can happen earlier (such as "too many connections") no longer contain
3008 12. If an expansion in a condition on a "warn" statement fails because a lookup
3009 defers, the "warn" statement is abandoned, and the next ACL statement is
3010 processed. Previously this caused the whole ACL to be aborted.
3012 13. Added the iplsearch lookup type.
3014 14. Added ident_timeout as a log selector.
3016 15. Added tls_certificate_verified as a log selector.
3018 16. Added a global option tls_require_ciphers (compare the smtp transport
3019 option of the same name). This controls incoming TLS connections.
3021 17. I finally figured out how to make tls_require_ciphers do a similar thing
3022 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3023 before starting the TLS session.
3025 18. Tabs are now shown as \t in -bP output.
3027 19. If the log selector return_path_on_delivery was set, Exim crashed when
3028 bouncing a message because it had too many Received: header lines.
3030 20. If two routers both had headers_remove settings, and the first one included
3031 a superfluous trailing colon, the final name in the first list and the
3032 first name in the second list were incorrectly joined into one item (with a
3033 colon in the middle).
3039 1. Added -C and -D options to the exinext utility, mainly to make it easier
3040 to include in the automated testing, but these could be helpful when
3041 multiple configurations are in use.
3043 2. The exinext utility was not formatting the output nicely when there was
3044 an alternate port involved in the retry record key, nor when there was a
3045 message id as well (for retries that were specific to a specific message
3046 and a specific host). It was also confused by IPv6 addresses, because of
3047 the additional colons they contain. I have fixed the IPv4 problem, and
3048 patched it up to do a reasonable job for IPv6.
3050 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3051 delivery, the log line now contains "pipelined" if PIPELINING was used.
3053 4. An SMTP transport process used to panic and die if the bind() call to set
3054 an explicit outgoing interface failed. This has been changed; it is now
3055 treated in the same way as a connect() failure.
3057 5. A reference to $sender_host_name in the part of a conditional expansion
3058 that was being skipped was still causing a DNS lookup. This no longer
3061 6. The def: expansion condition was not recognizing references to header lines
3062 that used bh_ and bheader_.
3064 7. Added the _cache feature to named lists.
3066 8. The code for checking quota_filecount in the appendfile transport was
3067 allowing one more file than it should have been.
3069 9. For compatibility with Sendmail, the command line option
3077 and sets the incoming protocol and host name (for trusted callers). The
3078 host name and its colon can be omitted when only the protocol is to be set.
3079 Note the Exim already has two private options, -pd and -ps, that refer to
3080 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3081 "s", but I don't think that's a major issue.
3083 10. A number of refactoring changes to the code, none of which should affect
3086 (a) The number of logging options was getting close to filling up the
3087 32-bit word that was used as a bit map. I have split them into two classes:
3088 those that are passed in the argument to log_write(), and those that are
3089 only ever tested independently outside of that function. These are now in
3090 separate 32-bit words, so there is plenty of room for expansion again.
3091 There is no change in the user interface or the logging behaviour.
3093 (b) When building, for example, log lines, the code previously used a
3094 macro that called string_cat() twice, in order to add two strings. This is
3095 not really sufficiently general. Furthermore, there was one instance where
3096 it was actually wrong because one of the argument was used twice, and in
3097 one call a function was used. (As it happened, calling the function twice
3098 did not affect the overall behaviour.) The macro has been replaced by a
3099 function that can join an arbitrary number of extra strings onto a growing
3102 (c) The code for expansion conditions now uses a table and a binary chop
3103 instead of a serial search (which was left over from when there were very
3104 few conditions). Also, it now recognizes conditions like "pam" even when
3105 the relevant support is not compiled in: a suitably worded error message is
3106 given if an attempt is made to use such a condition.
3108 11. Added ${time_interval:xxxxx}.
3110 12. A bug was causing one of the ddress fields not to be passed back correctly
3111 from remote delivery subprocesses. The field in question was not being
3112 subsequently used, so this caused to problems in practice.
3114 13. Added new log selectors queue_time and deliver_time.
3116 14. Might have fixed a bug in maildirsizefile handling that threw up
3117 "unexpected character" debug warnings, and recalculated the data
3118 unnecessarily. In any case, I expanded the warning message to give more
3121 15. Added the message "Restricted characters in address" to the statements in
3122 the default ACL that block characters like @ and % in local parts.
3124 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3125 Three changes have been made:
3127 (a) There was a serious bug; a negative response to MAIL caused the whole
3128 recipient domain to be cached as invalid, thereby blocking all messages
3129 to all local parts at the same domain, from all senders. This bug has
3130 been fixed. The domain is no longer cached after a negative response to
3131 MAIL if the sender used is not empty.
3133 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3136 (c) A new callout option, "use_sender" has been added for people who want
3137 the modified behaviour.
3143 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3146 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3147 indeed breaks things for older releases.
3149 3. Added additional logging to the case where there is a problem reading data
3150 from a filter that is running in a subprocess using a pipe, in order to
3151 try to track down a specific problem.
3153 4. Testing facility fudge: when running in the test harness and attempting
3154 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3155 getting "No route to host". Convert this to a timeout.
3157 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3160 6. Some OS don't have socklen_t but use size_t instead. This affects the
3161 fifth argument of getsockopt() amongst other things. This is now
3162 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3163 can be set for individual OS. I have set it for SunOS5, OSF1, and
3164 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3165 some earlier ones do not.
3167 7. Change 4.30/15 was not doing the test caselessly.
3169 8. The standard form for an IPv6 address literal was being rejected by address
3170 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3171 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3172 this, as well as the form without the "IPv6" on the front (but only when
3173 address literals are enabled, of course).
3175 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3177 10. Exim crashed if a message with an empty sender address specified by -f
3178 encountered a router with an errors_to setting. This could be provoked only
3179 by a command such as
3183 where an empty string was supplied; "<>" did not hit this bug.
3185 11. Installed PCRE release 4.5.
3187 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3188 remained set. It is now erased.
3190 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3191 times from message ids (which are base 36 rather than the normal 62).
3193 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3194 were being counted as actual protocol errors, and logged if the log
3195 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3196 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3197 and DATA following a set of rejected RCPTs do not count as protocol errors.
3198 In other words, Exim assumes they were pipelined, though this may not
3199 actually be the case. Of course, in all cases the client gets an
3200 appropriate error code.
3202 15. If a lookup fails in an ACL condition, a message about the failure may
3203 be available; it is used if testing the ACL cannot continue, because most
3204 such messages specify what the cause of the deferral is. However, some
3205 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3206 that caused an old message to be retained and used if a later statement
3207 caused a defer, replacing the real cause of the deferral.
3209 16. If an IP address had so many PTR records that the DNS lookup buffer
3210 was not large enough to hold them, Exim could crash while trying to process
3211 the truncated data. It now detects and logs this case.
3213 17. Further to 4.21/58, another change has been made: if (and only if) the
3214 first line of a message (the first header line) ends with CRLF, a bare LF
3215 in a subsequent header line has a space inserted after it, so as not to
3216 terminate the header.
3218 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3219 unnecessarily, used atoi() instead of strtol(), and didn't check the
3220 termination when getting file sizes from file names by regex.
3222 19. Completely re-implemented the support for maildirsize files, in the light
3223 of a number of problems with the previous contributed implementation
3224 (4.30/29). In particular:
3226 . If the quota is zero, the maildirsize file is maintained, but no quota is
3229 . If the maildir directory does not exist, it is created before any attempt
3230 to write a maildirsize file.
3232 . The quota value in the file is just a cache; if the quota is changed in
3233 the transport, the new value overrides.
3235 . A regular expression is available for excluding directories from the
3238 20. The autoreply transport checks the characters in options that define the
3239 message's headers; it allows continued headers, but it was checking with
3240 isspace() after an embedded newline instead of explicitly looking for a
3243 21. If all the "regular" hosts to which an address was routed had passed their
3244 expiry times, and had not reached their retry times, the address was
3245 bounced, even if fallback hosts were defined. Now Exim should go on to try
3248 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3249 equivalent code in the SMTP transport. Some hosts send humungous responses
3250 to HELO/EHLO, more than 1024 it seems.
3252 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3253 gives compiler warnings in some environments. I've now done it "properly",
3256 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3257 (because of problems with the built-in one) was declared to return uschar *
3258 instead of char *, causing compiler failure.
3260 25. Fixed a file descriptor leak when processing alias/forward files.
3262 26. Fixed a minor format string issue in dbfn.c.
3264 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3266 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3267 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3270 29. When a sender address is verified, it is cached, to save repeating the test
3271 when there is more than one recipient in a message. However, when the
3272 verification involves a callout, it is possible for different callout
3273 options to be set for different recipients. It is too complicated to keep
3274 track of this in the cache, so now Exim always runs a verification when a
3275 callout is required, relying on the callout cache for the optimization.
3276 The overhead is duplication of the address routing, but this should not be
3279 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3280 to be verified with callout=postmaster, and the main callout worked but the
3281 postmaster check failed, the verification correctly failed. However, if a
3282 subsequent RCPT command asked for sender verification *without* the
3283 postmaster check, incorrect caching caused this verification also to fail,
3286 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3287 it was not caching the DNS options (qualify_single, search_parents) that
3288 were used when the lookup failed. A subsequent lookup with different
3289 options therefore always gave the same answer, though there were cases
3290 where it should not have. (Example: a "domains = !$mx_any" option on a
3291 dnslookup router: the "domains" option is always processed without any
3292 widening, but the router might have qualify_single set.) Now Exim uses the
3293 cached value only when the same options are set.
3295 32. Added John Jetmore's "exipick" utility to the distribution.
3297 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3298 than a timeout (e.g. a certificate is required, and is not provided), an
3299 Exim server now closes the connection immediately. Previously it waited for
3300 the client to close - but if the client is SSL, it seems that they each
3301 wait for each other, leading to a delay before one of them times out.
3303 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3304 maintained 0.8.x compatibility because I don't think many are using it, and
3305 it is clearly obsolete.
3307 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3310 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3311 client certificate was expired. A simple patch fixes this, though I don't
3312 understand the full logic of why the verify callback is called multiple
3315 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3316 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3317 which causes problems with some clients (such as the Certicom SSL Plus
3318 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3319 disables the coutermeasure allowing Eudora to connect."
3321 38. Exim was not checking that a write() to a log file succeeded. This could
3322 lead to Bad Things if a log got too big, in particular if it hit a file
3323 size limit. Exim now panics and dies if it cannot write to a log file, just
3324 as it does if it cannot open a log file.
3326 39. Modified OS/Makefile-Linux so that it now contains
3328 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3330 The two -D definitions ensure that Exim is compiled with large file
3331 support, which makes it possible to handle log files that are bigger than
3334 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3335 instance) a domain was checked against a named list that involved a lookup,
3336 causing $domain_data to be set, then another domain was checked against the
3337 same list, then the first domain was re-checked, the value of $domain_data
3338 after the final check could be wrong. In particular, if the second check
3339 failed, it could be set empty. This bug probably also applied to
3342 41. The strip_trailing_dot option was not being applied to the address given
3343 with the -f command-line option.
3345 42. The code for reading a message's header from the spool was incrementing
3346 $received_count, but never initializing it. This meant that the value was
3347 incorrect (doubled) while delivering a message in the same process in which
3348 it was received. In the most common configuration of Exim, this never
3349 happens - a fresh exec is done - but it can happen when
3350 deliver_drop_privilege is set.
3352 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3353 it now includes up to 150 characters of the unexpected data in the log
3356 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3357 and building data strings. The size of both of these buffers was 10 000
3358 bytes - far larger than anybody would *ever* want, thought I. Needless to
3359 say, somebody hit the limit. I have increased the maximum line length to
3360 20 000 and the maximum data length of concatenated lines to 100 000. I have
3361 also fixed two bugs, because there was no checking on these buffers. Tsk,
3362 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3363 buffer is too small.
3365 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3366 lsearch lookups. Now it does.
3368 46. When parsing a route_list item in a manualroute router, a fixed-length
3369 buffer was used for the list of hosts. I made this 1024 bytes long,
3370 thinking that nobody would ever have a list of hosts that long. Wrong.
3371 Somebody had a whole pile of complicated expansion conditions, and the
3372 string was silently truncated, leading to an expansion error. It turns out
3373 that it is easier to change to an unlimited length (owing to other changes
3374 that have happened since this code was originally written) than to build
3375 structure for giving a limitation error. The length of the item that
3376 expands into the list of hosts is now unlimited.
3378 47. The lsearch lookup could not handle data where the length of text line was
3379 more than 4095 characters. Such lines were truncated, leading to shortened
3380 data being returned. It should now handle lines of any length.
3382 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3383 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3386 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3387 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3388 no longer happens in comment lines. A list of replacements is now placed
3389 at the head of all of the source files, except those whose only change is
3390 to replace PERL_COMMAND in the very first #! line.
3392 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3393 messages on the queue, with a bottom-up merge sort, using code contributed
3394 by Michael Haardt. This should make operations like -bp somewhat faster on
3395 large queues. It won't affect queue runners, except when queue_run_in_order
3398 51. Installed eximstats 1.31 in the distribution.
3400 52. Added support for SRV lookups to the dnslookup router.
3402 53. If an ACL referred to $message_body or $message_body_end, the value was not
3403 reset for any messages that followed in the same SMTP session.
3405 54. The store-handling optimization for building very long strings was not
3406 differentiating between the different store pools. I don't think this
3407 actually made any difference in practice, but I've tidied it.
3409 55. While running the routers to verify a sender address, $sender_address
3410 was still set to the sender address. This is wrong, because when routing to
3411 send a bounce to the sender, it would be empty. Therefore, I have changed
3412 it so that, while verifying a sender address, $sender_address is set to <>.
3413 (There is no change to what happens when verifying a recipient address.)
3415 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3416 target A or AAAA records (if not already returned) without resetting the
3417 qualify_single or search_parents options of the DNS resolver. These are
3418 inappropriate in this case because the targets of MX and SRV records must
3419 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3420 target that, when qualified, matched something in the local domain. These
3421 two options are now turned off when doing these lookups.
3423 57. It seems that at least some releases of Reiserfs (which does not have the
3424 concept of a fixed number of inodes) returns zero and not -1 for the
3425 number of available inodes. This interacted badly with check_spool_inodes,
3426 which assumed that -1 was the "no such thing" setting. What I have done is
3427 to check that the total number of inodes is greater than zero before doing
3428 the test of how many are available.
3430 58. When a "warn" ACL statement has a log_message modifier, the message is
3431 remembered, and not repeated. This is to avoid a lot of repetition when a
3432 message has many recipients that cause the same warning to be written.
3433 Howewer, Exim was preserving the list of already written lines for an
3434 entire SMTP session, which doesn't seem right. The memory is now reset if a
3435 new message is started.
3437 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3438 debugging output unless log_rewrite was also set.
3440 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3441 of (int)(handle) when we know that handle contains (void *)(-1).
3443 61. The Exim daemon panic-logs an error return when it closes the incoming
3444 connection. However "connection reset by peer" seems to be common, and
3445 isn't really an error worthy of noting specially, so that particular error
3448 62. When Exim is trying to find all the local interfaces, it used to panic and
3449 die if the ioctl to get the interface flags failed. However, it seems that
3450 on at least one OS (Solaris 9) it is possible to have an interface that is
3451 included in the list of interfaces, but for which you get a failure error
3452 for this call. This happens when the interface is not "plumbed" into a
3453 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3454 failure of the "get flags" call assumes that the interface is down.
3456 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3457 makes life easier for people who are doing arithmetic on fields extracted
3458 from dates, where you often get leading zeros that should not be
3459 interpreted as octal.
3461 64. Added qualify_domain to the redirect router, to override the global
3464 65. If a pathologically long header line contained very many addresses (the
3465 report of this problem mentioned 10 000) and each of them was rewritten,
3466 Exim could use up a very large amount of memory. (It kept on making new
3467 copies of the header line as it rewrote, and never released the old ones.)
3468 At the expense of a bit more processing, the header rewriting function has
3469 been changed so that it no longer eats memory in this way.
3471 66. The generation of the Received: header has been moved from the time that a
3472 message starts to be received, to the time that it finishes. The timestamp
3473 in the Received: header should now be very close to that of the <= log
3474 line. There are two side-effects of this change:
3476 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3477 logged header lines no longer include the local Received: line, because
3478 it has not yet been created. The same applies to a copy of the message
3479 that is returned to a non-SMTP sender when a message is rejected.
3481 (b) When a filter file is tested using -bf, no additional Received: header
3482 is added to the test message. After some thought, I decided that this
3485 This change does not affect the value of $received_for. It is still set
3486 after address rewriting, but before local_scan() is called.
3488 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3490 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3491 gave an unhelpful panic error message, and a defer error. I have managed to
3492 change this behaviour so that it now rejects any supplied certificate,
3493 which seems right, as the list of acceptable certificates is empty.
3495 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3496 gave an unhelpful defer error. I have not managed to make this reject any
3497 supplied certificates, but the error message it gives is "no certificate
3498 supplied", which is not helpful.
3500 70. exigrep's output now also includes lines that are not associated with any
3501 message, but which match the given pattern. Implemented by a patch from
3502 Martin Sluka, which also tidied up the Perl a bit.
3504 71. Recipient callout verification, like sender verification, was using <> in
3505 the MAIL FROM command. This isn't really the right thing, since the actual
3506 sender may affect whether the remote host accepts the recipient or not. I
3507 have changed it to use the actual sender in the callout; this means that
3508 the cache record is now keyed on a recipient/sender pair, not just the
3509 recipient address. There doesn't seem to be a real danger of callout loops,
3510 since a callout by the remote host to check the sender would use <>.
3511 [SEE ABOVE: changed after hitting problems.]
3513 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3514 temporary errors. However, in the case of such a code being given after
3515 the end of a data transmission (i.e. after ".") Exim was failing to write
3516 a retry record for the message. (Yes, there was some broken host that was
3517 actually sending 8xx at this point.)
3519 73. An unknown lookup type in a host list could cause Exim to panic-die when
3520 the list was checked. (An example that provoked this was putting <; in the
3521 middle of a list instead of at the start.) If this happened during a DATA
3522 ACL check, a -D file could be left lying around. This kind of configuration
3523 error no longer causes Exim to die; instead it causes a defer errror. The
3524 incident is still logged to the main and panic logs.
3526 74. Buglet left over from Exim 3 conversion. The message "too many messages
3527 in one connection" was written to the rejectlog but not the mainlog, except
3528 when address rewriting (yes!) was being logged.
3530 75. Added write_rejectlog option.
3532 76. When a system filter was run not as root (that is, when system_filter_user
3533 was set), the values of the $n variables were not being returned to the
3534 main process; thus, they were not subsequently available in the $sn
3537 77. Added +return_path_on_delivery log selector.
3539 78. A connection timeout was being treated differently from recipients deferred
3540 when testing hosts_max_try with a message that was older than the host's
3541 retry timeout. (The host should not be counted, thus allowing all hosts to
3542 be tried at least once before bouncing.) This may have been the cause of an
3543 occasionally reported bug whereby a message would remain on the queue
3544 longer than the retry timeout, but would be bounced if a delivery was
3545 forced. I say "may" because I never totally pinned down the problem;
3546 setting up timeout/retry tests is difficult. See also the next item.
3548 79. The ultimate address timeout was not being applied to errors that involved
3549 a combination of host plus message (for example, a timeout on a MAIL
3550 command). When an address resolved to a number of possible hosts, and they
3551 were not all tried for each delivery (e.g. because of hosts_max_try), a
3552 message could remain on the queue longer than the retry timeout.
3554 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3557 81. Fixed an obscure SMTP outgoing bug which required at least the following
3558 conditions: (a) there was another message waiting for the same server;
3559 (b) the server returned 5xx to all RCPT commands in the first message so
3560 that the message was not completed; (c) the server dropped the connection
3561 or gave a negative response to the RSET that Exim sends to abort the
3562 transaction. The observed case was a dropped connection after DATA that had
3563 been sent in pipelining mode. That is, the server had advertised PIPELINING
3564 but was not implementing it correctly. The effect of the bug was incorrect
3565 behaviour, such as trying another host, and this could lead to a crash.
3571 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3572 and daemon.c were passed as pointers to ints; they should have been
3573 pointers to socklen_t variables (which are typically unsigned ints).
3575 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3578 3. Fixed a really odd bug that affected only the testing scheme; patching a
3579 certain fixed string in the binary changed the value of another string that
3580 happened to be identical to the end of the original first string.
3582 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3583 name", it returns that address as the IP address. On some operating
3584 systems (e.g. Solaris), it also passes back the IP address string as the
3585 "host name". However, on others (e.g. Linux), it passes back an empty
3586 string. Exim wasn't checking for this, and was changing the host name to an
3587 empty string, assuming it had been canonicized.
3589 5. Although rare, it is permitted to have more than one PTR record for a given
3590 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3591 all the names associated with an address, because they do in Solaris.
3592 However, it seems that they do not in Linux for data that comes from the
3593 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3594 I found this out when I moved to a new Linux workstation and tried to run
3595 the Exim test suite.
3597 To get round this problem I have changed the code so that it now does its
3598 own call to the DNS to look up PTR records when searching for a host name.
3599 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3600 addresses that are only in /etc/hosts are still found.
3602 This behaviour is, however, controlled by an option called host_lookup_
3603 order, which defaults to "bydns:byaddr". If people want to use the other
3604 order, or indeed, just use one or the other means of lookup, they can
3605 specify it in this variable.
3607 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3608 some operating systems, this comes back from gethostbyaddr() as an empty
3609 string, and this is what Exim used to test for. However, it seems that in
3610 other systems, "." is yielded. Exim now tests for this case too.
3612 7. The values of check_spool_space and check_log_space are now held internally
3613 as a number of kilobytes instead of an absolute number of bytes. If a
3614 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3615 kilobyte. This means that much larger values can be stored.
3617 8. Exim monitor: an attempt to get the action menu when not actually pointing
3618 at a message produces an empty menu entitled "No message selected". This
3619 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3620 no entries in it ("Shell widget menu has zero width and/or height"). So I
3621 have added a single, blank menu entry in this case.
3623 9. Added ${quote_local_part.
3625 10. MIME decoding is now applied to the contents of Subject: header lines when
3628 11. Now that a reference to $sender_host_address automatically causes a reverse
3629 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3630 host lookup before query-style lookups in lists that might use this
3631 variable. This has therefore been abolished, and the "net-" prefix is no
3632 longer necessary for query-style lookups.
3634 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3635 have been a typo for LFLAGS, so it has been changed.
3637 13. The install script calls Exim with "-C /dev/null" in order to find the
3638 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3639 to be output. Howeve, since Exim outputs its version number before the
3640 error, it didn't break the script. It just looked ugly. I fixed this by
3641 always allowing "-C /dev/null" if the caller is root.
3643 14. Ignore overlarge ACL variable number when reading spool file - insurance
3644 against a later release with more variables having written the file.
3646 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3647 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3648 this, as well as the form without the "IPv6" on the front.
3650 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3651 OS/Makefile-Darwin file.
3653 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3654 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3656 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3657 variable. However, some OS use ints for inodes. I've added cast to long int
3658 to get rid of the compiler warning.
3660 19. I had forgotten to lock out "/../" in configuration file names when
3661 ALT_CONFIG_PREFIX was set.
3663 20. Routers used for verification do not need to specify transports. However,
3664 if such a router generated a host list, and callout was configured, Exim
3665 crashed, because it could not find a port number from the (non-existent)
3666 transport. It now assumes port 25 in this circumstance.
3668 21. Added the -t option to exigrep.
3670 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3671 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3672 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3673 them is set, LOOKUP_LSEARCH is forced.
3675 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3677 24. Added sender and host information to the "rejected by local_scan()" log
3678 line; previously there was no indication of these.
3680 25. Added .include_if_exists.
3682 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3683 Linux. It turns out that not all file systems support this. Apparently some
3684 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3685 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3686 not supported on the file descriptor, is now ignored when Exim is trying to
3687 sync a directory. This applies only to Linux.
3689 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3691 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3692 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3695 29. Added support for maildirsize files from supplied patch (modified a bit).
3697 30. The use of :fail: followed by an empty string could lead Exim to respond to
3698 sender verification failures with (e.g.):
3700 550 Verification failed for <xxx>
3701 550 Sender verify failed
3703 where the first response line was missing the '-' that indicates it is not
3704 the final line of the response.
3706 31. The loop for finding the name of the user that called Exim had a hardwired
3707 limit of 10; it now uses the value of finduser_retries, which is used for
3708 all other user lookups.
3710 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3713 33. Exim was neglecting to zero errno before one call of strtol() when
3714 expanding a string and expecting an integer value. On some systems this
3715 resulted in spurious "integer overflow" errors. Also, it was casting the
3716 result into an int without checking.
3718 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3719 did not work. The code looks as if it has *never* worked, though it appears
3720 to have been documented since at least releast 1.62. I have made it work.
3722 35. The "timeout_DNS" error in retry rules, also documented since at least
3723 1.62, also never worked. As it isn't clear exactly what this means, and
3724 clearly it isn't a major issue, I have abolished the feature by treating it
3725 as "timeout", and writing a warning to the main and panic logs.
3727 36. The display of retry rules for -brt wasn't always showing the error code
3730 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3731 timeout_connect_A, timeout_connect_MX.
3733 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3734 to the empty sender.
3736 39. The daemon was not analysing the content of -oX till after it had closed
3737 stderr and disconnected from the controlling terminal. This meant that any
3738 syntax errors were only noted on the panic log, and the return code from
3739 the command was 0. By re-arranging the code a little, I've made the
3740 decoding happen first, so such errors now appear on stderr, and the return
3741 code is 1. However, the actual setting up of the sockets still happens in
3742 the disconnected process, so errors there are still only recorded on the
3745 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3746 connections (as happens on some IP stacks) was logged at start up time as
3747 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3748 it from "IPv6 and IPv4", which means that two separate sockets are being
3751 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3752 says whether AF_INET or AF_INET6 was passed as an argument.
3754 42. Exiwhat output was messed up when time zones were included in log
3757 43. Exiwhat now gives more information about the daemon's listening ports,
3758 and whether -tls-on-connect was used.
3760 44. The "port" option of the smtp transport is now expanded.
3762 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3763 silently ignored. Now an error message is written to the main and panic
3766 46. There's a new ACL modifier called "logwrite" which writes to a log file
3767 as soon as it is encountered.
3769 47. Added $local_user_uid and $local_user_gid at routing time.
3771 48. Exim crashed when trying to verify a sender address that was being
3774 49. Exim was recognizing only a space character after ".include". It now also
3775 recognizes a tab character.
3777 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3778 extracting the relevant information from the specification. The man page no
3779 longer contains scrambled data for the -d option, and I've added a section
3780 at the front about calling Exim under different names.
3782 51. Added "extra_headers" argument to the "mail" command in filter files.
3784 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3787 53. Installed eximstats 1.29.
3789 54. Added transport_filter_timeout as a generic transport option.
3791 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3792 Cc: header lines. This was required by RFC 822, but it not required by RFC
3795 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3796 incoming messages that did not have them. Now it does so only if the
3797 message originates locally, that is, if there is no associated remote host
3798 address. When Resent- header lines are present, this applies to the Resent-
3799 lines rather than the non-Resent- lines.
3801 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3802 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3804 58. Messages for configuration errors now include the name of the main
3805 configuration file - useful now that there may be more than one file in a
3806 list (.included file names were always shown).
3808 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3809 for those rare installations that do not start the daemon as root or run it
3810 setuid root. I've cut out the call to initgroups() if the daemon is not
3813 60. The Exim user and group can now be bound into the binary as text strings
3814 that are looked up at the start of Exim's processing.
3816 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3818 62. Added $mailstore_basename variable.
3820 63. Installed patch to sieve.c from Michael Haardt.
3822 64. When Exim failed to open the panic log after failing to open the main log,
3823 the original message it was trying to log was written to stderr and debug
3824 output, but if they were not available (the usual case in production), it
3825 was lost. Now it is written to syslog before the two lines that record the
3826 failures to open the logs.
3828 65. Users' Exim filters run in subprocesses under the user's uid. It is
3829 possible for a "deliver" command or an alias in a "personal" command to
3830 provoke an address rewrite. If logging of address rewriting is configured,
3831 this fails because the process is not running as root or exim. There may be
3832 a better way of dealing with this, but for the moment (because 4.30 needs
3833 to be released), I have disabled address rewrite logging when running a
3834 filter in a non-root, non-exim process.
3840 1. The buildconfig auxiliary program wasn't quoting the value set for
3841 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3842 not defined. This bug was masked in 4.22 by the effect that was fixed in
3845 2. Some messages that were rejected after a message id was allocated were
3846 shown as "incomplete" by exigrep. It no longer does this for messages that
3847 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3849 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3850 have allow_domain_literals set, the ID did not get logged in the <= line.
3851 Domain literals are now always recognized in Message-ID: header lines.
3853 4. The first argument for a ${extract expansion item is the key name or field
3854 number. Leading and trailing spaces in this item were not being ignored,
3855 causing some misleading effects.
3857 5. When deliver_drop_privilege was set, single queue runner processes started
3858 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3859 same command in the process it spins off) were not dropping privilege.
3861 6. When the daemon running as "exim" started a queue runner, it always
3862 re-executed Exim in the spun-off process. This is a waste of effort when
3863 deliver_drop_privilege is set. The new process now just calls the
3864 queue-runner function directly.
3870 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3873 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3874 ignored. Though the use of -oP was forcing the writing of a pid file, it
3875 was always written to the default place.
3877 3. If the message "no IP address found for host xxxx" is generated during
3878 incoming verification, it is now followed by identification of the incoming
3879 connection (so you can more easily find what provoked it).
3881 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3883 5. Added some features to "harden" Exim a bit more against certain attacks:
3885 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3886 be put in Local/Makefile. This is like the never_users runtime option,
3887 but it cannot be overridden. The default setting is "root".
3889 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3890 prefix string with which any file named in a -C command line option
3893 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3894 is retained for -C and -D only if the caller of Exim is root. Without
3895 it, the exim user may also use -C and -D and retain privilege.
3897 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3898 command line option is disabled.
3900 6. Macro names set by the -D option must start with an upper case letter, just
3901 like macro names defined in the configuration file.
3903 7. Added "dereference=" facility to LDAP.
3905 8. Two instances of the typo "uknown" in the source files are fixed.
3907 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3908 the Configure-Makefile script screwed up while processing it.
3910 10. Incorporated PCRE 4.4.
3912 11. The SMTP synchronization check was not operating right at the start of an
3913 SMTP session. For example, it could not catch a HELO sent before the client
3914 waited for the greeting. There is now a check for outstanding input at the
3915 point when the greeting is written. Because of the duplex, asynchronous
3916 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3917 way, but not yet received, when the check is performed.
3919 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3920 on TCP/IP sockets, because this apparently causes some broken clients to
3923 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3924 unchanged) from the Cygwin maintainer.
3926 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3927 was supported instead of testing for mbx. Effectively a typo.
3929 15. The spa authenticator server code was not checking that the input it
3930 received was valid base64.
3932 16. The debug output line for the "set" modifier in ACLs was not showing the
3933 name of the variable that was being set.
3935 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3937 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3938 lookup on its own. Something else was needed to trigger the lookup. For
3939 example, a match in host_lookup or the need for a host name in a host list.
3940 Now, if $sender_host_name is referenced and the host name has not yet been
3941 looked up, a lookup is performed. If the lookup fails, the variable remains
3942 empty, and $host_lookup_failed is set to "1".
3944 19. Added "eqi" as a case-independent comparison operator.
3946 20. The saslauthd authentication condition could segfault if neither service
3947 nor realm was specified.
3949 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3950 error message that was logged was misleading, and incoming SMTP
3951 connections were dropped. The message is now more accurate, and temporary
3952 errors are given to SMTP connections.
3954 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3955 (except RSET) with a 421 error, until QUIT is received. However, it was
3956 failing to send a response to QUIT.
3958 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3959 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3960 ACL is now run after the helo_try_verify_hosts code.
3962 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3963 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3964 case-independent, so other case variants are also recognized). Apparently
3965 some systems use these upper case variants.
3967 25. If more than two messages were waiting for the same host, and a transport
3968 filter was specified for the transport, Exim sent two messages over the
3969 same TCP/IP connection, and then failed with "socket operation on non-
3970 socket" when it tried to send the third.
3972 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3974 27. The extern definition of crypt16() in expand.c was not being excluded when
3975 the OS had its own crypt16() function.
3977 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3978 as a preferred synonym for return_size_limit, both as an option and as an
3981 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3983 30. Changed the default configuration ACL to relax the local part checking rule
3984 for addresses that are not in any local domains. For these addresses,
3985 slashes and pipe symbols are allowed within local parts, but the sequence
3986 /../ is explicitly forbidden.
3988 31. SPA server authentication was not clearing the challenge buffer before
3991 32. log_message in a "warn" ACL statement was writing to the reject log as
3992 well as to the main log, which contradicts the documentation and doesn't
3993 seem right (because no rejection is happening). So I have stopped it.
3995 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3996 However, I am unable to do any testing of this.
3998 34. Fixed an infelicity in the appendfile transport. When checking directories
3999 for a mailbox, to see if any needed to be created, it was accidentally
4000 using path names with one or more superfluous leading slashes; tracing
4001 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4003 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4004 contents are added to the log line that is written for every discarded
4005 recipient. (Previously a log_message setting was ignored.)
4007 36. The ${quote: operator now quotes the string if it is empty.
4009 37. The install script runs exim in order to find its version number. If for
4010 some reason other than non-existence or emptiness, which it checks, it
4011 could not run './exim', it was installing it with an empty version number,
4012 i.e. as "exim-". This error state is now caught, and the installation is
4015 38. An argument was missing from the function that creates an error message
4016 when Exim fails to connect to the socket for saslauthd authentication.
4017 This could cause Exim to crash, or give a corrupted message.
4019 39. Added isip, isip4, and isip6 to ${if conditions.
4021 40. The ACL variables $acl_xx are now saved with the message, and can be
4022 accessed later in routers, transports, and filters.
4024 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4025 strings in the file are not string-expanded.
4027 42. If a MAIL command specified a SIZE value that was too large to fit into an
4028 int variable, the check against message_size_limit failed. Such values are
4029 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4030 day this will have to be increased, but I don't think I want to be around
4031 when emails are that large.
4038 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4039 iconv() is not standard in FreeBSD.
4041 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4042 IPv6 enabled. The observed symptom was a segmentation fault on return from
4043 the function os_common_find_running_interfaces() in src/os.c.
4045 3. In the check_special_case() function in daemon.c I had used "errno" as an
4046 argument name, which causes warnings on some systems. This was basically a
4047 typo, since it was named "eno" in the comments!
4049 4. The code that waits for the clock to tick (at a resolution of some fraction
4050 of a second) so as to ensure message-id uniqueness was always waiting for
4051 at least one whole tick, when it could have waited for less. [This is
4052 almost certainly not relevant at current processor speeds, where it is
4053 unlikely to ever wait at all. But we try to future-proof.]
4055 5. The function that sleeps for a time interval that includes fractions of a
4056 second contained a race. It did not block SIGALRM between setting the
4057 timer, and suspending (a couple of lines later). If the interval was short
4058 and the sigsuspend() was delayed until after it had expired, the suspension
4059 never ended. On busy systems this could lead to processes getting stuck for
4062 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4063 process, before it forks any delivery processes. The open lookup caching
4064 mechanism meant that the open file or database connection was passed into
4065 the delivery process. The problem was that delivery processes always tidy
4066 up cached lookup data. This could cause a problem for the next delivery
4067 process started by the queue runner, because the external queue runner
4068 process does not know about the closure. So the next delivery process
4069 still has data in the lookup cache. In the case of a file lookup, there was
4070 no problem because closing a file descriptor in a subprocess doesn't affect
4071 the parent. However, if the lookup was caching a connection to a database,
4072 the connection was closed, and the second delivery process was likely to
4073 see errors such as "PGSQL: query failed: server closed the connection
4074 unexpectedly". The problem has been fixed by closing all cached lookups
4075 in a queue runner before running a delivery process.
4077 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4078 seem to have the "const" qualifier which it has on other OS. I've
4081 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4082 *of the same type* (client or server) with the same public name that an
4083 error should be diagnosed.
4085 9. When Exim looked up a host name for an IP address, but failed to find the
4086 original IP address when looking up the host name (a safety check), it
4087 output the message "<ip address> does not match any IP for NULL", which was
4088 confusing, to say the least. The bug was that the host name should have
4089 appeared instead of "NULL".
4091 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4092 user that is built into the binary, and the -C or -D options is used, root
4093 privilege is dropped before the configuration file is read. In addition,
4094 logging is switched to stderr instead of the normal log files. If the
4095 configuration then re-defines the Exim user, the unprivileged environment
4096 is probably not what is expected, so Exim logs a panic warning message (but
4099 However, if deliver_drop_privilege is set, the unprivileged state may well
4100 be exactly what is intended, so the warning has been cut out in that case,
4101 and Exim is allowed to try to write to its normal log files.
4107 1. smtp_return_error_details was not giving details for temporary sender
4108 or receiver verification errors.
4110 2. Diagnose a configuration error if two authenticators have the same public
4113 3. Exim used not to create the message log file for a message until the first
4114 delivery attempt. This could be confusing when incoming messages were held
4115 for policy or load reasons. The message log file is now created at the time
4116 the message is received, and an initial "Received" line is written to it.
4118 4. The automatically generated man page for command line options had a minor
4119 bug that caused no ill effects; however, a more serious problem was that
4120 the procedure for building the man page automatically didn't always
4121 operate. Consequently, release 4.20 contains an out-of-date version. This
4122 shouldn't happen again.
4124 5. When building Exim with embedded Perl support, the script that builds the
4125 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4126 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4128 6. The freeze_tell option was not being used for messages that were frozen on
4129 arrival, either by an ACL or by local_scan().
4131 7. Added the smtp_incomplete_transaction log selector.
4133 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4134 was accepting AUTH without a new EHLO.
4136 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4137 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4140 10. Logging of TCP/IP connections (when configured) now happens in the main
4141 daemon process instead of the child process, so that the TCP/IP connection
4142 count is more accurate (but it can never be perfect).
4144 11. The use of "drop" in a nested ACL was not being handled correctly in the
4145 outer ACL. Now, if condition failure induced by the nested "drop" causes
4146 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4147 or "require"), the connection is dropped.
4149 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4150 that yield "discard" can now be used with an "accept" or a "discard" verb,
4151 but an error is generated for any others (because I can't see a useful way
4152 to define what should happen).
4154 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4155 are now processed in the same way as lines in the Exim configuration file.
4156 In particular, continuation lines are supported.
4158 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4160 15. Added -ti meaning -t -i.
4162 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4163 domains, and warn by logging if others are found.
4165 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4166 returned from GIFCONF when Exim is trying to find the list of interfaces on
4167 a host. The code in os.c has been modified to copy each ifreq to an aligned
4168 structure in all cases.
4170 Also, in some cases, the returned ifreq's were being copied to a 'struct
4171 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4172 means the last couple of bytes of an IPv6 address could be chopped if the
4173 ifreq contained only a normal sockaddr (14 bytes storage).
4175 18. Named domain lists were not supported in the hosts_treat_as_local option.
4176 An entry such as +xxxx was not recognized, and was treated as a literal
4179 19. Ensure that header lines added by a DATA ACL are included in the reject log
4180 if the ACL subsequently rejects the message.
4182 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4183 MD5 (which is deprecated).
4185 21. When testing a filter file using -bf, Exim was writing a message when it
4186 took the sender from a "From " line in the message, but it was not doing so
4187 when it took $return_path from a Return-Path: header line. It now does.
4189 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4190 with a valid header line field name (a series of printing characters
4191 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4193 23. Changed "disc" in the source to "disk" to conform to the documentation and
4194 the book and for uniformity.
4196 24. Ignore Sendmail's -Ooption=value command line item.
4198 25. When execve() failed while trying to run a command in a pipe transport,
4199 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4200 could be confused with a return value of 69 from the command itself. This
4201 has been changed to 127, the value the shell returns if it is asked to run
4202 a non-existent command. The wording for the related log line suggests a
4203 non-existent command as the problem.
4205 26. If received_header_text expands to an empty string, do not add a Received:
4206 header line to the message. (Well, it adds a token one on the spool, but
4207 marks it "old" so that it doesn't get used or transmitted.)
4209 27. Installed eximstats 1.28 (addition of -nt option).
4211 28. There was no check for failure on the call to getsockname() in the daemon
4212 code. This can fail if there is a shortage of resources on the system, with
4213 ENOMEM, for example. A temporary error is now given on failure.
4215 29. Contrary to the C standard, it seems that in some environments, the
4216 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4217 program. Exim now does this explicitly; it affects the formatting of
4218 timestamps using strftime().
4220 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4221 complaints. I've now initialized all the variables, to avoid this.
4223 32. Header lines added by a system filter were not being "seen" during
4224 transport-time rewrites.
4226 33. The info_callback() function passed to OpenSSL is set up with type void
4227 (*)(SSL *, int, int), as described somewhere. However, when calling the
4228 function (actually a macro) that sets it up, the type void(*)() is
4229 expected. I've put in a cast to prevent warnings from picky compilers.
4231 34. If a DNS black list lookup found a CNAME record, but there were no A
4232 records associated with the domain it pointed at, Exim crashed.
4234 35. If a DNS black list lookup returned more than one A record, Exim ignored
4235 all but the first. It now scans all returned addresses if a particular IP
4236 value is being sought. In this situation, the contents of the
4237 $dnslist_value variable are a list of all the addresses, separated by a
4240 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4241 to accept a host name and all its aliases if the forward lookup for any of
4242 them yielded the IP address of the incoming connection. Now it accepts only
4243 those names whose forward lookup yields the correct IP address. Any other
4244 names are discarded. This closes a loophole whereby a rogue DNS
4245 administrator could create reverse DNS records to break through a
4246 wildcarded host restriction in an ACL.
4248 37. If a user filter or a system filter that ran in a subprocess used any of
4249 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4250 the wrong values were passed to the pipe command ($thisaddress had the
4251 value of $0, $0 had the value of $1, etc). This bug was introduced by
4252 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4254 38. Improved the line breaking for long SMTP error messages from ACLs.
4255 Previously, if there was no break point between 40 and 75 characters, Exim
4256 left the rest of the message alone. Two changes have been made: (a) I've
4257 reduced the minimum length to 35 characters; (b) if it can't find a break
4258 point between 35 and 75 characters, it looks ahead and uses the first one
4259 that it finds. This may give the occasional overlong line, but at least the
4260 remaining text gets split now.
4262 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4263 file descriptors might be low, and that setting 1000 would always raise it.
4264 It turns out that in some environments, the limit is already over 1000 and
4265 that lowering it causes trouble. So now Exim takes care not to decrease it.
4267 40. When delivering a message, the value of $return_path is set to $sender_
4268 address at the start of routing (routers may change the value). By an
4269 oversight, this default was not being set up when an address was tested by
4270 -bt or -bv, which affected the outcome if any router or filter referred to
4273 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4274 log only when "message" or "log_message" are set. However, if one of the
4275 conditions was an address verification, or a call to a nested ACL, the
4276 messages generated by the underlying test were being passed through. This
4277 no longer happens. The underlying message is available in $acl_verify_
4278 message for both "message" and "log_message" expansions, so it can be
4279 passed through if needed.
4281 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4282 new expansion $bh_ to give the encoded byte string without charset
4283 translation. Translation happens only if iconv() is available; HAVE_ICONV
4284 indicates this at build time. HEADERS_CHARSET gives the charset to
4285 translate to; headers_charset can change it in the configuration, and
4286 "headers charset" can change it in an individual filter file.
4288 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4289 that creates RFC 2047 encoded "words" labels them as that charset instead
4290 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4291 expansion operator; (ii) when Exim creates a From: line for a local
4292 message; (iii) when a header line is rewritten to include a "phrase" part.
4294 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4295 buggy, causing it to skip the first lines of messages whose message ID
4296 ended in 'D'. This would not have bitten before Exim release 4.14, because
4297 message IDs were unlikely to end in 'D' before then. The effect was to have
4298 incorrect size information for certain domains.
4300 45. #include "config.h" was missing at the start of the crypt16.c module. This
4301 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4304 46. If there was a timeout during a "random" callout check, Exim treated it as
4305 a failure of the random address, and carried on sending RSET and the real
4306 address. If the delay was just some slowness somewhere, the response to the
4307 original RCPT would be taken as a response to RSET and so on, causing
4308 mayhem of various kinds.
4310 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4311 when I implemented it. It didn't allow for the fact that some option values
4312 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4313 the right test for positive values.
4315 48. Domain names in DNS records are case-independent. Exim always looks them up
4316 in lower case. Some resolvers return domain names in exactly the case they
4317 appear in the zone file, that is, they may contain uppercase letters. Not
4318 all resolvers do this - some return always lower case. Exim was treating a
4319 change of case by a resolver as a change of domain, similar to a widening
4320 of a domain abbreviation. This triggered its re-routing code and so it was
4321 trying to route what was effectively the same domain again. This normally
4322 caused routing to fail (because the router wouldn't handle the domain
4323 twice). Now Exim checks for this case specially, and just changes the
4324 casing of the domain that it ultimately uses when it transmits the message
4327 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4330 50. If a filter generated a file delivery with a non-absolute name (possible if
4331 no home directory exists for the router), the forbid_file option was not
4334 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4335 the existing equality matching.
4337 52. Exim was using ints instead of ino_t variables in some places where it was
4338 dealing with inode numbers.
4340 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4341 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4342 called TMPDIR, and if it finds it is different, it changes its value.
4344 54. The smtp_printf() function is now made available to local_scan() so
4345 additional output lines can be written before returning. There is also an
4346 smtp_fflush() function to enable the detection of a dropped connection.
4347 The variables smtp_input and smtp_batched_input are exported to
4350 55. Changed the default runtime configuration: the message "Unknown user"
4351 has been removed from the ACL, and instead placed on the localuser router,
4352 using the cannot_route_message feature. This means that any verification
4353 failures that generate their own messages won't get overridden. Similarly,
4354 the "Unrouteable address" message that was in the ACL for unverifiable
4355 relay addresses has also been removed.
4357 56. Added hosts_avoid_esmtp to the smtp transport.
4359 57. The exicyclog script was not checking for the esoteric option
4360 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4361 will work only if exicyclog is run under the appropriate euid.
4363 58. Following a discussion on the list, the rules by which Exim recognises line
4364 endings on incoming messages have been changed. The -dropcr and drop_cr
4365 options are now no-ops, retained only for backwards compatibility. The
4366 following line terminators are recognized: LF CRLF CR. However, special
4367 processing applies to CR:
4369 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4370 nor a local message in the state where . is a terminator.
4372 (ii) If a bare CR is encountered in a header line, an extra space is added
4373 after the line terminator so as not to end the header. The reasoning
4374 behind this is that bare CRs in header lines are most likely either
4375 to be mistakes, or people trying to play silly games.
4377 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4378 was being incorrectly given as 18 bytes larger than it should have been.
4379 This is a VOB (very old bug).
4381 60. This may never have affected anything current, but just in case it has:
4382 When the local host is found other than at the start of a list of hosts,
4383 the local host, those with the same MX, and any that follow, are discarded.
4384 When the list in question was part of a longer list of hosts, the following
4385 hosts (not currently being processed) were also being discarded. This no
4386 longer happens. I'm not sure if this situation could ever has previously
4389 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4392 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4393 header. This is something that is recommended in a new Internet Draft, and
4394 is something that is documented as being done by Sendmail. There are two
4395 possible values. For messages generated by the autoreply transport, Exim
4398 Auto-Submitted: auto-replied
4400 whereas for all other generated messages (e.g. bounces) it adds
4402 Auto-Submitted: auto-generated
4404 63. The "personal" condition in filters now includes a test for the
4405 Auto-Submitted: header. If it contains the string "auto-" the message it
4406 not considered personal.
4408 64. Added rcpt_include_affixes as a generic transport option.
4410 65. Added queue_only_override (default true).
4412 66. Added the syslog_duplication option.
4414 67. If what should have been the first header line of a message consisted of
4415 a space followed by a colon, Exim was mis-interpreting it as a header line.
4416 It isn't of course - it is syntactically invalid and should therefore be
4417 treated as the start of the message body. The misbehaviour could have
4418 caused a number of strange effects, including loss of data in subsequent
4419 header lines, and spool format errors.
4421 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4422 client host had authenticated. This control can now be exercised by an ACL
4423 for more flexibility.
4425 69. By default, callouts do not happen when testing with -bh. There is now a
4426 variant, -bhc, which does actually run the callout code, including
4427 consulting and updating the callout cache.
4429 70. Added support for saslauthd authentication, courtesy of Alexander
4432 71. If statvfs() failed on the spool or log directories while checking their
4433 size for availability, Exim confusingly gave the error "space shortage".
4434 Furthermore, in debugging mode it crashed with a floating point exception.
4435 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4436 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4437 serious problem, Exim now writes to the main and panic logs when this
4438 happens, with details of the failure. It then refuses to accept the
4439 incoming message, giving the message "spool directory problem" or "log
4440 directory problem" with a 421 code for SMTP messages.
4442 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4443 0, 1, and 2 exist, because some OS complain for execs without them (see
4444 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4445 descriptors. However, the code omitted to check that the open succeeded,
4446 causing mysterious errors if for some reason the permissions on /dev/null
4447 got screwed. Now Exim writes a message to the main and panic logs, and
4448 bombs out if it can't open /dev/null.
4450 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4451 interact so that it is all more flexible. It is supposed to remain
4452 backwards compatible. Also added extra_local_interfaces.
4454 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4455 to bomb out with an assertion failure - to the client this appears as a
4456 connection drop. This problem occurs in the part of the code that was taken
4457 from the Samba project. Fortunately, the assertion is in a very simple
4458 function, so I have fixed this by reproducing the function inline in the
4459 one place where it is called, and arranging for authentication to fail
4460 instead of killing the process with assert().
4462 75. The SPA client code was not working when the server requested OEM rather
4463 than Unicode encoding.
4465 76. Added code to make require_files with a specific uid setting more usable in
4466 the case where statting the file as root fails - usually a non-root-mounted
4467 NFS file system. When this happens and the failure is EACCES, Exim now
4468 forks a subprocess and does the per-uid checking as the relevant uid.
4470 77. Added process_log_path.
4472 78. If log_file_path was not explicitly set, a setting of check_log_space or
4473 check_log_inodes was ignored.
4475 79. If a space check for the spool or log partitions fails, the incident is now
4476 logged. Of course, in the latter case the data may get lost...
4478 80. Added the %p formatting code to string_format() so that it can be used to
4479 print addresses in debug_print(). Adjusted all the address printing in the
4480 debugging in store.c to use %p rather than %d.
4482 81. There was a concern that a line of code in smtp_in.c could overflow a
4483 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4484 initially expressed, the concern was not well-founded, because trailing
4485 spaces are removed early. However, if the trailing spaces were followed by
4486 a NULL, they did not get removed, so the overflow was possible. Two fixes
4489 (a) I re-wrote the offending code in a cleaner fashion.
4490 (b) If an incoming SMTP command contains a NULL character, it is rejected
4493 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4494 runs initgroups(), so that if the Exim user is in any additional groups,
4495 they will be used during message reception.
4501 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob