1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.462 2007/01/23 15:08:45 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
61 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
62 fixed by 4.65/MH/01 (is this a record?) are fixed:
64 (i) An empty string was always treated as zero by the numeric comparison
65 operators. This behaviour has been restored.
67 (ii) It is documented that the numeric comparison operators always treat
68 their arguments as decimal numbers. This was broken in that numbers
69 starting with 0 were being interpreted as octal.
71 While fixing these problems I realized that there was another issue that
72 hadn't been noticed. Values of message_size_limit (both the global option
73 and the transport option) were treated as octal if they started with 0.
74 The documentation was vague. These values are now always treated as
75 decimal, and I will make that clear in the documentation.
81 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
82 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
85 MH/01 Don't check that the operands of numeric comparison operators are
86 integers when their expansion is in "skipping" mode (fixes bug
87 introduced by 4.64-PH/07).
89 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
90 child addresses, Exim now panics and dies. Previously, because the count
91 is held in a short int, deliveries were likely to be lost. As such a
92 large number of recipients for a single message is ridiculous
93 (performance will be very, very poor), I have chosen to impose a limit
94 rather than extend the field.
100 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
101 leftover -K file (the existence of which was triggered by #402).
102 While we were at it, introduced process PID as part of the -K
103 filename. This should rule out race conditions when creating
106 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
107 processing considerably. Previous code took too long for large mails,
108 triggering a timeout which in turn triggers #401.
110 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
111 in the DK code in transports.c. sendfile() is not really portable,
112 hence the _LINUX specificness.
114 TF/01 In the add_headers option to the mail command in an Exim filter,
115 there was a bug that Exim would claim a syntax error in any
116 header after the first one which had an odd number of characters
119 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
120 callout verification, Exim cached a "reject" for the entire domain. This
121 is correct for most verifications, but it is not correct for a recipient
122 verification with use_sender or use_postmaster set, because in that case
123 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
124 case of MAIL FROM:<> rejection from other early rejections (e.g.
125 rejection of HELO). When verifying a recipient using a non-null MAIL
126 address, the cache is ignored if it shows MAIL FROM:<> rejection.
127 Whatever the result of the callout, the value of the domain cache is
128 left unchanged (for any other kind of callout, getting as far as trying
129 RCPT means that the domain itself is ok).
131 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
134 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
135 manifest itself as EPIPE rather than ECONNECT. When tidying away a
136 session, the daemon ignores ECONNECT errors and logs others; it now
137 ignores EPIPE as well.
139 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
140 (quoted-printable decoding).
142 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
143 later the small subsequent patch to fix an introduced bug.
145 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
147 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
149 PH/08 An error is now given if message_size_limit is specified negative.
151 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
152 to be given (somewhat) arbitrary names.
154 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
157 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
158 miscellaneous code fixes
160 PH/10 Added the log_reject_target ACL modifier to specify where to log
163 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
164 hostname. This is wrong, because it relates to the incoming message (and
165 probably the interface on which it is arriving) and not to the outgoing
166 callout (which could be using a different interface). This has been
167 changed to use the value of the helo_data option from the smtp transport
168 instead - this is what is used when a message is actually being sent. If
169 there is no remote transport (possible with a router that sets up host
170 addresses), $smtp_active_hostname is used.
172 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
173 tweaks were necessary in order to get it to work (see also 21 below):
174 (a) The code assumed that strncpy() returns a negative number on buffer
175 overflow, which isn't the case. Replaced with Exim's string_format()
177 (b) There were several signed/unsigned issues. I just did the minimum
178 hacking in of casts. There is scope for a larger refactoring.
179 (c) The code used strcasecmp() which is not a standard C function.
180 Replaced with Exim's strcmpic() function.
181 (d) The code set only $1; it now sets $auth1 as well.
182 (e) A simple test gave the error "authentication client didn't specify
183 service in request". It would seem that Dovecot has changed its
184 interface. Fortunately there's a specification; I followed it and
185 changed what the client sends and it appears to be working now.
187 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
190 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
191 address A is aliased to B and C, where B exists and C does not. Without
192 -v the output is "A verified" because verification stops after a
193 successful redirection if more than one address is generated. However,
194 with -v the child addresses are also verified. Exim was outputting "A
195 failed to verify" and then showing the successful verification for C,
196 with its parentage. It now outputs "B failed to verify", showing B's
197 parentage before showing the successful verification of C.
199 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
200 look up a TXT record in a specific list after matching in a combined
203 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
204 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
205 they consult the DNS. I had assumed they would set it the way they
206 wanted; and indeed my experiments on Linux seem to show that in some
207 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
208 To be on the safe side, however, I have now made the interface to
209 host_find_byname() similar to host_find_bydns(), with an argument
210 containing the DNS resolver options. The host_find_byname() function now
211 sets these options at its start, just as host_find_bydns() does. The smtp
212 transport options dns_qualify_single and dns_search_parents are passed to
213 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
214 of host_find_byname() use the default settings of RES_DEFNAMES
215 (qualify_single) but not RES_DNSRCH (search_parents).
217 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
218 spool_read_header() do less string testing, by means of a preliminary
219 switch on the second character of optional "-foo" lines. (This is
220 overdue, caused by the large number of possibilities that now exist.
221 Originally there were few.) While I was there, I also converted the
222 str(n)cmp tests so they don't re-test the leading "-" and the first
223 character, in the hope this might squeeze out yet more improvement.
225 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
226 flag allowing group syntax was set by the header_syntax check but not
227 turned off, possible causing trouble later; (2) The flag was not being
228 set at all for the header_verify test, causing "group"-style headers to
229 be rejected. I have now set it in this case, and also caused header_
230 verify to ignore an empty address taken from a group. While doing this, I
231 came across some other cases where the code for allowing group syntax
232 while scanning a header line wasn't quite right (mostly, not resetting
233 the flag correctly in the right place). These bugs could have caused
234 trouble for malformed header lines. I hope it is now all correct.
236 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
237 with the "reply" argument non-NULL. The code, however (which originally
238 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
239 but it didn't always do it. This confused somebody who was copying the
240 code for some other use. I have removed all the tests.
242 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
243 feature that was used to support insecure browsers during the U.S. crypto
244 embargo. It requires special client support, and Exim is probably the
245 only MTA that supported it -- and would never use it because real RSA is
246 always available. This code has been removed, because it had the bad
247 effect of slowing Exim down by computing (never used) parameters for the
248 RSA_EXPORT functionality.
250 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
251 authenticator to fail if there's a tab character in the incoming data
252 (there should never be unless someone is messing about, as it's supposed
253 to be base64-encoded). Also added, on Timo's advice, the "secured" option
254 if the connection is using TLS or if the remote IP is the same as the
255 local IP, and the "valid-client-cert option" if a client certificate has
258 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
259 authenticators. This can be used for authorization after authentication
260 succeeds. (In the case of plaintext, it servers for both authentication
263 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
264 if any retry times were supplied.
266 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
267 connection, where there is no client IP address to check. In this
268 situation, the verify now always succeeds.
270 PH/25 Applied John Jetmore's -Mset patch.
272 PH/26 Added -bem to be like -Mset, but loading a message from a file.
274 PH/27 In a string expansion for a processed (not raw) header when multiple
275 headers of the same name were present, leading whitespace was being
276 removed from all of them, but trailing whitespace was being removed only
277 from the last one. Now trailing whitespace is removed from each header
278 before concatenation. Completely empty headers in a concatenation (as
281 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
282 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
284 PH/29 [Removed. This was a change that I later backed out, and forgot to
285 correct the ChangeLog entry (that I had efficiently created) before
286 committing the later change.]
288 PH/30 Exim was sometimes attempting to deliver messages that had suffered
289 address errors (4xx response to RCPT) over the same connection as other
290 messages routed to the same hosts. Such deliveries are always "forced",
291 so retry times are not inspected. This resulted in far too many retries
292 for the affected addresses. The effect occurred only when there were more
293 hosts than the hosts_max_try setting in the smtp transport when it had
294 the 4xx errors. Those hosts that it had tried were not added to the list
295 of hosts for which the message was waiting, so if all were tried, there
296 was no problem. Two fixes have been applied:
298 (i) If there are any address or message errors in an SMTP delivery, none
299 of the hosts (tried or untried) are now added to the list of hosts
300 for which the message is waiting, so the message should not be a
301 candidate for sending over the same connection that was used for a
302 successful delivery of some other message. This seems entirely
303 reasonable: after all the message is NOT "waiting for some host".
304 This is so "obvious" that I'm not sure why it wasn't done
305 previously. Hope I haven't missed anything, but it can't do any
306 harm, as the worst effect is to miss an optimization.
308 (ii) If, despite (i), such a delivery is accidentally attempted, the
309 routing retry time is respected, so at least it doesn't keep
310 hammering the server.
312 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
313 in ${readsocket because some servers need this prod.
315 PH/32 Added some extra debug output when updating a wait-xxx database.
317 PH/33 The hint "could be header name not terminated by colon", which has been
318 given for certain expansion errors for a long time, was not being given
319 for the ${if def:h_colon_omitted{... case.
321 PH/34 The spec says: "With one important exception, whenever a domain list is
322 being scanned, $domain contains the subject domain." There was at least
323 one case where this was not true.
325 PH/35 The error "getsockname() failed: connection reset by peer" was being
326 written to the panic log as well as the main log, but it isn't really
327 panic-worthy as it just means the connection died rather early on. I have
328 removed the panic log writing for the ECONNRESET error when getsockname()
331 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
332 runs only) independently of the message's sender address. This meant
333 that, if the 4xx error was in fact related to the sender, a different
334 message to the same recipient with a different sender could confuse
335 things. In particualar, this can happen when sending to a greylisting
336 server, but other circumstances could also provoke similar problems.
337 I have changed the default so that the retry time for these errors is now
338 based a combination of the sender and recipient addresses. This change
339 can be overridden by setting address_retry_include_sender=false in the
342 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
343 remote server are returned as part of bounce messages. This was not
344 happening for LMTP over a pipe (the lmtp transport), but now it is the
345 same for both kinds of LMTP.
347 PH/38 Despite being documented as not happening, Exim was rewriting addresses
348 in header lines that were in fact CNAMEs. This is no longer the case.
350 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
351 and queue runs started by the daemon processed all messages. This has
352 been fixed so that -R and -S can now usefully be given with -q<time>.
354 PH/40 Import PCRE release 6.7 (fixes some bugs).
356 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
358 PH/42 Give an error if -q is specified more than once.
360 PH/43 Renamed the variables $interface_address and $interface_port as
361 $received_ip_address and $received_port, to make it clear that these
362 values apply to message reception, and not to the outgoing interface when
363 a message is delivered. (The old names remain recognized, of course.)
365 PH/44 There was no timeout on the connect() call when using a Unix domain
366 socket in the ${readsocket expansion. There now is.
368 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
369 be meaningful with "accept".
371 SC/01 Eximstats V1.43
372 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
374 SC/02 Eximstats V1.44
375 Use a glob alias rather than an array ref in the generated
376 parser. This improves both readability and performance.
378 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
379 Collect SpamAssassin and rejection statistics.
380 Don't display local sender or destination tables unless
381 there is data to show.
382 Added average volumes into the top table text output.
384 SC/04 Eximstats V1.46
385 Collect data on the number of addresses (recipients)
386 as well as the number of messages.
388 SC/05 Eximstats V1.47
389 Added 'Message too big' to the list of mail rejection
390 reasons (thanks to Marco Gaiarin).
392 SC/06 Eximstats V1.48
393 Mainlog lines which have GMT offsets and are too short to
394 have a flag are now skipped.
396 SC/07 Eximstats V1.49 (Alain Williams)
397 Added the -emptyok flag.
399 SC/08 Eximstats V1.50
400 Fixes for obtaining the IP address from reject messages.
402 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
403 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
404 whitesspace changes from 4.64-PH/27
406 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
409 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
410 are found, allow negative numbers in numeric criteria)
412 JJ/06 exipick.20061117.2, added new $message_body_missing variable
414 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
415 to match changes made in 4.64-PH/43
417 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
419 PH/47 Put in an explicit test for a DNS lookup of an address record where the
420 "domain" is actually an IP address, and force a failure. This locks out
421 those revolvers/nameservers that support "A-for-A" lookups, in
422 contravention of the specifications.
424 PH/48 When a host name was looked up from an IP address, and the subsequent
425 forward lookup of the name timed out, the host name was left in
426 $sender_host_name, contrary to the specification.
428 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
429 restricted to single-key lookups, Exim was not diagnosing an error if
430 * or *@ was used with a query-style lookup.
432 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
434 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
435 long ago, but noone interested enough thought of it. Let's just say that
436 the "1.1" means that there are some new functions that weren't there at
437 some point in the past.
439 PH/51 Error processing for expansion failure of helo_data from an smtp
440 transport during callout processing was broken.
442 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
443 tested/used via the -bh/-bhc/-bs options.
445 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
446 bug, fixed in subsequent PCRE releases).
448 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
449 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
451 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
457 SC/01 Use a glob alias rather than an array ref in eximstats generated
458 parser. This improves both readability and performance.
460 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
461 Don't display local sender or destination tables in eximstats unless
462 there is data to show.
463 Added average volumes into the eximstats top table text output.
465 SC/03 Collect data on the number of addresses (recipients) as well
466 as the number of messages in eximstats.
468 TF/01 Correct an error in the documentation for the redirect router. Exim
469 does (usually) call initgroups() when daemonizing.
471 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
472 with consistent privilege compared to when running as a daemon.
474 TF/03 Note in the spec that $authenticated_id is not set for local
475 submissions from trusted users.
477 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
478 Thanks to Dean Brooks <dean@iglou.com> for the patch.
480 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
481 by adding some example configuration directives to the default
482 configuration file. A little bit of work is required to uncomment the
483 directives and define how usernames and passwords are checked, but
484 there is now a framework to start from.
486 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
487 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
488 without this. I don't know how relevant this is to other LDAP libraries.
490 PH/02 Add the verb name to the "unknown ACL verb" error.
492 PH/03 Magnus Holmgren's patch for filter_prepend_home.
494 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
496 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
497 directory not expanded when it should be if an expanded home directory
498 was set for the address (which is overridden by the transport).
500 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
503 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
504 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
505 because it is too late at that time, and has no effect.
507 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
508 security issue with \' (bugzilla #107). I could not use the
509 PQescapeStringConn() function, because it needs a PGconn value as one of
512 PH/08 When testing addresses using -bt, indicate those final addresses that
513 are duplicates that would not cause an additional delivery. At least one
514 person was confused, thinking that -bt output corresponded to deliveries.
515 (Suppressing duplicates isn't a good idea as you lose the information
516 about possibly different redirections that led to the duplicates.)
518 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
519 systems where poll() doesn't work, in particular OS X.
521 PH/10 Added more information to debugging output for retry time not reached.
523 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
524 operations in malware.c.
526 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
529 PH/13 If write_rejectlog was set false when logging was sent to syslog with
530 syslog_duplication set false, log lines that would normally be written
531 both the the main log and to the reject log were not written to syslog at
534 PH/14 In the default configuration, change the use of "message" in ACL warn
535 statements to "add_header".
537 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
538 not followed by a command (e.g. "seen endif").
540 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
541 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
544 PH/17 Added extra conditions to the default value of delay_warning_condition
548 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
549 { match{$h_precedence:}{(?i)bulk|list|junk} } \
550 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
553 The Auto-Submitted: and various List- headers are standardised, whereas I
554 don't think Precedence: ever was.
556 PH/18 Refactored debugging code in route_finduser() to show more information,
557 in particular, the error code if getpwnam() issues one.
559 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
560 This is apparently needed in addition to the PH/07 change above to avoid
561 any possible encoding problems.
563 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
564 but not after initializing Perl.
566 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
567 output them only if debugging. By default they are written stderr,
568 apparently, which is not desirable.
570 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
573 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
576 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
578 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
579 authenticated or an ident call has been made. Suppress the default
580 values for $authenticated_id and $authenticated_sender (but permit -oMai
581 and -oMas) when testing with -bh.
583 PH/24 Re-jigged the order of the tests in the default configuration so that the
584 tests for valid domains and recipients precede the DNS black list and CSA
585 tests, on the grounds that those ones are more expensive.
587 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
588 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
589 This bug exists in every version of Exim that I still have, right back to
592 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
593 However, an attempt to turn on case-sensitivity in a regex key by
594 including (?-i) didn't work because the subject string was already
595 lowercased, and the effects were non-intuitive. It turns out that a
596 one-line patch can be used to allow (?-i) to work as expected.
602 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
603 other effects) broke the use of negated acl sub-conditions.
605 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
608 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
609 "Deny" causes Exim to reject the incoming connection with a 554 error.
610 Unfortunately, if there is a major crisis, such as a disk failure,
611 tcp-wrappers gives "deny", whereas what one would like would be some
612 kind of temporary error. A kludge has been added to help with this.
613 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
614 554 error is used if errno is still zero or contains ENOENT (which occurs
615 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
618 PH/03 Add -lutil to the default FreeBSD LIBS setting.
620 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
621 errors. Otherwise a message that provokes a temporary error (when other
622 messages do not) can cause a whole host to time out.
624 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
625 addresses were routed directly to files or pipes from a redirect router.
626 File deliveries just didn't batch; pipe deliveries might have suffered
629 PH/06 A failure to get a lock for a hints database would erroneously always say
630 "Failed to get write lock", even when it was really a read lock.
632 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
633 of 0600. This has been changed to use the value of the lockfile_mode
634 option (which defaults to 0600).
636 PH/08 Applied small patch from the Sieve maintainer.
638 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
639 folder from quota calculations, a direct delivery into this folder messed
640 up the contents of the maildirsize file. This was because the regex was
641 used only to exclude .Trash (or whatever) when the size of the mailbox
642 was calculated. There was no check that a delivery was happening into an
643 excluded directory. This bug has been fixed by ignoring all quota
644 processing for deliveries into excluded directories.
646 PH/10 Added the maildirfolder_create_regex option to appendfile.
652 PH/01 The code for finding all the local interface addresses on a FreeBSD
653 system running IPv6 was broken. This may well have applied to all BSD
654 systems, as well as to others that have similar system calls. The broken
655 code found IPv4 interfaces correctly, but gave incorrect values for the
656 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
657 that it would not match correctly against @[] and not recognize the IPv6
660 PH/02 The ipliteral router was not recognizing addresses of the form user@
661 [ipv6:....] because it didn't know about the "ipv6:" prefix.
663 PH/03 Added disable_ipv6.
665 PH/04 Changed $reply_address to use the raw form of the headers instead of the
666 decoded form, because it is most often used to construct To: headers
667 lines in autoreplies, and the decoded form may well be syntactically
668 invalid. However, $reply_address has leading white space removed, and all
669 newlines turned into spaces so that the autoreply transport does not
672 PH/05 If group was specified without a user on a router, and no group or user
673 was specified on a transport, the group from the router was ignored.
675 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
676 for visible compile-time settings that can be used to change these
677 numbers, for those that want even more. Backwards compatibility with old
678 spool files has been maintained. However, going back to a previous Exim
679 release will lost any variables that are in spool files.
681 PH/07 Two small changes when running in the test harness: increase delay when
682 passing a TCP/IP connection to a new process, in case the original
683 process has to generate a bounce, and remove special handling of
684 127.0.0.2 (sic), which is no longer necessary.
686 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
687 be the same on different OS.
689 PH/09 Moved a debug statement in filter processing to avoid a race problem when
692 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
693 whether --show-vars was specified or not
695 JJ/02 exipick: Added support for new ACL variable spool format introduced
698 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
699 syntactically invalid From: or Reply-to: line, and a filter used this to
700 generate an autoreply, and therefore failed to obtain an address for the
701 autoreply, Exim could try to deliver to a non-existent relative file
702 name, causing unrelated and misleading errors. What now happens is that
703 it logs this as a hard delivery error, but does not attempt to create a
706 PH/11 The exinext utility has a -C option for testing purposes, but although
707 the given file was scanned by exinext itself; it wasn't being passed on
708 when Exim was called.
710 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
711 an end-of-file indication when reading a command response.
713 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
714 compiled. In many other places in Exim, IPv6 addresses are always
715 recognized, so I have changed this. It also means that IPv4 domain
716 literals of the form [IPV4:n.n.n.n] are now always recognized.
718 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
719 used if the router is not running as root, for example, when verifying at
720 ACL time, or when using -bh. The debugging output from this situation was
721 non-existent - all you got was a failure to exec. I have made two
724 (a) Failures to set uid/gid, the current directory, or a process leader
725 in a subprocess such as that created by queryprogram now generate
726 suitable debugging ouput when -d is set.
728 (b) The queryprogram router detects when it is not running as root,
729 outputs suitable debugging information if -d is set, and then runs
730 the subprocess without attempting to change uid/gid.
732 PH/15 Minor change to Makefile for building test_host (undocumented testing
735 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
736 additional section of a DNS packet that returns MX or SRV records.
737 Instead, it always explicitly searches for A/AAAA records. This avoids
738 major problems that occur when a DNS server includes only records of one
739 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
740 fixed another bug: if SRV records were looked up and the corresponding
741 address records were *not* found in the additional section, the port
742 values from the SRV records were lost.
744 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
745 using the correct key (the original address) when searching the retry
746 rules in order to find which one to use for generating the retry hint.
748 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
749 adding the default one. Similarly, if it contains a Reply-To: header, the
750 errors_reply_to option, if set, is not used.
752 PH/19 When calculating a retry time, Exim used to measure the "time since
753 failure" by looking at the "first failed" field in the retry record. Now
754 it does not use this if it is later than than the arrival time of the
755 message. Instead it uses the arrival time. This makes for better
756 behaviour in cases where some deliveries succeed, thus re-setting the
757 "first failed" field. An example is a quota failure for a huge message
758 when small messages continue to be delivered. Without this change, the
759 "time since failure" will always be short, possible causing more frequent
760 delivery attempts for the huge message than are intended.
761 [Note: This change was subsequently modified - see PH/04 for 4.62.]
763 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
764 $1, $2, $3) because the numerical variables can be reset during some
765 expansion items (e.g. "match"), thereby losing the authentication data.
767 PH/21 Make -bV show the size of off_t variables so that the test suite can
768 decide whether to run tests for quotas > 2G.
770 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
771 mailbox_size, and mailbox_filecount in the appendfile transport. If a
772 filecount value is greater than 2G or if a quota value is greater than 2G
773 on a system where the size of off_t is not greater than 4, a panic error
776 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
777 never match. The debug and -bh output now contains an explicit error
778 message indicating a malformed IPv4 address or mask.
780 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
781 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
784 PH/25 Do not write to syslog when running in the test harness. The only
785 occasion when this arises is a failure to open the main or panic logs
786 (for which there is an explicit test).
788 PH/26 Added the /no_tell option to "control=freeze".
790 PH/27 If a host name lookup failed very early in a connection, for example, if
791 the IP address matched host_lookup and the reverse lookup yielded a name
792 that did not have a forward lookup, an error message of the form "no IP
793 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
794 could be logged. Now it outputs the IP address instead of "NULL".
796 PH/28 An enabling patch from MH: add new function child_open_exim2() which
797 allows the sender and the authenticated sender to be set when
798 submitting a message from within Exim. Since child_open_exim() is
799 documented for local_scan(), the new function should be too.
801 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
802 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
803 results in an empty string is now treated as unset.
805 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
807 PH/31 Added sender_verify_fail logging option.
809 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
810 needed by RFC 822 but not by RFC 2822 was commented out. I have now
811 tidied the source and removed it altogether.
813 PH/33 When a queue run was abandoned because the load average was too high, a
814 log line was always written; now it is written only if the queue_run log
815 selector is set. In addition, the log line for abandonment now contains
816 information about the queue run such as the pid. This is always present
817 in "start" and "stop" lines but was omitted from the "abandon" line.
819 PH/34 Omit spaces between a header name and the colon in the error message that
820 is given when verify = headers_syntax fails (if there are lots of them,
821 the message gets confusing).
823 PH/35 Change the default for dns_check_names_pattern to allow slashes within
824 names, as there are now some PTR records that contain slashes. This check
825 is only to protect against broken name servers that fall over on strange
826 characters, so the fact that it applies to all lookups doesn't matter.
828 PH/36 Now that the new test suite is complete, we can remove some of the
829 special code in Exim that was needed for the old test suite. For example,
830 sorting DNS records because real resolvers return them in an arbitrary
831 order. The new test suite's fake resolver always returns records in the
834 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
835 bounces) except when queue_only is set, to avoid logging races between
836 the different processes.
838 PH/38 Panic-die if .include specifies a non-absolute path.
840 PH/39 A tweak to the "H" retry rule from its user.
842 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
843 a label. They prevented compilation on older perls.
845 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
846 a warning to be raised on newish perls.
848 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
849 on queue. Changes to match documented behaviour of showing count of
850 messages matching specified criteria.
852 PH/40 Changed the default ident timeout from 30s to 5s.
854 PH/41 Added support for the use of login_cap features, on those BSD systems
855 that have them, for controlling the resources used by pipe deliveries.
857 PH/42 The content-scanning code uses fopen() to create files in which to put
858 message data. Previously it was not paying any attention to the mode of
859 the files. Exim runs with umask(0) because the rest of the code creates
860 files with open(), and sets the required mode explicitly. Thus, these
861 files were ending up world-writeable. This was not a big issue, because,
862 being within the spool directory, they were not world-accessible. I have
863 created a function called modefopen, which takes an additional mode
864 argument. It sets umask(777), creates the file, chmods it to the required
865 mode, then resets the umask. All the relevant calls to fopen() in the
866 content scanning code have been changed to use this function.
868 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
869 to 24 hours. This avoids potential overflow problems when processing G
870 and H retry rules. I suspect nobody ever tinkers with this value.
872 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
874 PH/45 When the plaintext authenticator is running as a client, the server's
875 challenges are checked to ensure they are valid base64 strings. By
876 default, the authentication attempt is cancelled if an invalid string is
877 received. Setting client_ignore_invalid_base64 true ignores these errors.
878 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
879 they are received. Thus, the responses can be made to depend on the
880 challenges. If an invalid string is ignored, an empty string is placed in
883 PH/46 Messages that are created by the autoreply transport now contains a
884 References: header, in accordance with RFCs 2822 and 3834.
886 PH/47 Added authenticated_sender_force to the smtp transport.
888 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
890 PH/49 Installed latest patch from the Sieve maintainer.
892 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
893 was also set, the appendfile transport was unnecessarily scanning a
894 directory of message files (e.g. for maildir delivery) to find the count
895 of files (along with the size), even though it did not need this
896 information. It now does the scan only if it needs to find either the
897 size of the count of files.
899 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
901 PH/52 Two bugs concerned with error handling when the smtp transport is
904 (i) Exim was not creating retry information for temporary errors given
905 for individual recipients after the DATA command when the smtp transport
906 was used in LMTP mode. This meant that they could be retried too
907 frequently, and not timed out correctly.
909 (ii) Exim was setting the flag that allows error details to be returned
910 for LMTP errors on RCPT commands, but not for LMTP errors for individual
911 recipients that were returned after the DATA command.
913 PH/53 This is related to PH/52, but is more general: for any failing address,
914 when detailed error information was permitted to be returned to the
915 sender, but the error was temporary, then after the final timeout, only
916 "retry timeout exceeded" was returned. Now it returns the full error as
917 well as "retry timeout exceeded".
919 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
920 do this, and (what is worse) MTAs that accept it.
922 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
923 will now be deprecated.
925 PH/56 New os.c-cygwin from the Cygwin maintainer.
927 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
928 formats (previously only available in exim formats via -bpr, -bpru,
929 and -bpra. Now also available in native and exiqgrep formats)
931 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
932 with very large, slow to parse queues
934 JJ/08 exipick: added ! as generic prefix to negate any criteria format
936 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
938 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
939 responses to authentication challenges, though it was showing the
940 challenges; (ii) I've removed the CR characters from the debug output for
943 PH/58 Allow for the insertion of a newline as well as a space when a string
944 is turned into more than one encoded-word during RFC 2047 encoding. The
945 Sieve code now uses this.
947 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
948 data_4xx, lost_connection, tls_required.
950 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
951 message was being sent as an SMTP response.
953 PH/61 Add -l and -k options to exicyclog.
955 PH/62 When verifying, if an address was redirected to one new address, so that
956 verification continued, and the new address failed or deferred after
957 having set something in $address_data, the value of $address_data was not
958 passed back to the ACL. This was different to the case when no
959 redirection occurred. The value is now passed back in both cases.
961 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
962 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
963 use login_cap.h, so on its own it isn't the distinguishing feature. The
964 new name refers directly to the setclassresources() function.
966 PH/65 Added configuration files for NetBSD3.
968 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
970 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
971 is preferred over IPv4.
973 PH/68 The bounce_return_message and bounce_return_body options were not being
974 honoured for bounces generated during the reception of non-SMTP messages.
975 In particular, this applied to messages rejected by the ACL. This bug has
976 been fixed. However, if bounce_return_message is true and bounce_return_
977 body is false, the headers that are returned for a non-SMTP message
978 include only those that have been read before the error was detected.
979 (In the case of an ACL rejection, they have all been read.)
981 PH/69 The HTML version of the specification is now built in a directory called
982 spec_html instead of spec.html, because the latter looks like a path with
983 a MIME-type, and this confuses some software.
985 PH/70 Catch two compiler warnings in sieve.c.
987 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
988 function verify_get_ident() calls ip_connect() to connect a socket, but
989 if the "connect()" function timed out, ip_connect() used to close the
990 socket. However, verify_get_ident() also closes the socket later, and in
991 between Exim writes to the log, which may get opened at this point. When
992 the socket was closed in ip_connect(), the log could get the same file
993 descriptor number as the socket. This naturally causes chaos. The fix is
994 not to close the socket in ip_connect(); the socket should be closed by
995 the function that creates it. There was only one place in the code where
996 this was missing, in the iplookup router, which I don't think anybody now
997 uses, but I've fixed it anyway.
999 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1000 well as to direct DNS lookups. Otherwise the handling of names in host
1001 lists is inconsistent and therefore confusing.
1007 PH/01 Two changes to the default runtime configuration:
1009 (1) Move the checks for relay_from_hosts and authenticated clients from
1010 after to before the (commented out) DNS black list checks.
1012 (2) Add control=submission to the relay_from_hosts and authenticated
1013 clients checks, on the grounds that messages accepted by these
1014 statements are most likely to be submissions.
1016 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1018 (1) Generate an error if the third argument for the ${prvs expansion is
1021 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1024 (3) Reset the variables that are obtained from the first argument of
1025 ${prvscheck and used in the second argument before leaving the code,
1026 because their memory is reclaimed, so using them afterwards may do
1029 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1030 one (it's much easier than Tom thought :-).
1032 (5) Because of (4), we can now allow for the use of $prvscheck_result
1033 inside the third argument.
1035 PH/03 For some reason, the default setting of PATH when running a command from
1036 a pipe transport was just "/usr/bin". I have changed it to
1039 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1040 anything to be listed in the output from -bV.
1042 PH/05 When a filter generated an autoreply, the entire To: header line was
1043 quoted in the delivery log line, like this:
1045 => >A.N.Other <ano@some.domain> <original@ddress> ...
1047 This has been changed so that it extracts the operative address. There
1048 may be more than one such address. If so, they are comma-separated, like
1051 => >ano@some.domain,ona@other.domain <original@ddress> ...
1053 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1054 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1055 not being looked up in the rDNS to get a host name, Exim was showing the
1056 IP address twice in Received: lines, even though the IP addresses were
1057 identical. For example:
1059 Received: from [1.2.3.4] (helo=[1.2.3.4])
1061 However, if the real host name was known, it was omitting the HELO data
1062 if it matched the actual IP address. This has been tidied up so that it
1063 doesn't show the same IP address twice.
1065 PH/07 When both +timestamp and +memory debugging was on, the value given by
1066 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1067 called by the debug printing, thereby overwriting the timestamp buffer.
1068 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1070 PH/08 When the original message was included in an autoreply transport, it
1071 always said "this is a copy of the message, including all the headers",
1072 even if body_only or headers_only was set. It now gives an appropriate
1075 PH/09 Applied a patch from the Sieve maintainer which:
1077 o fixes some comments
1078 o adds the (disabled) notify extension core
1079 o adds some debug output for the result of if/elsif tests
1080 o points to the current vacation draft in the documentation
1081 and documents the missing references header update
1085 o fixes a bug in processing the envelope test (when testing
1086 multiple envelope elements, the last element determinted the
1089 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1090 Electronic Mail") by including:
1092 Auto-submitted: auto-generated
1094 in the messages that it generates (bounce messages and others, such as
1095 warnings). In the case of bounce messages for non-SMTP mesages, there was
1096 also a typo: it was using "Auto_submitted" (underscore instead of
1097 hyphen). Since every message generated by Exim is necessarily in response
1098 to another message, thes have all been changed to:
1100 Auto-Submitted: auto-replied
1102 in accordance with these statements in the RFC:
1104 The auto-replied keyword:
1106 - SHOULD be used on messages sent in direct response to another
1107 message by an automatic process,
1109 - MUST NOT be used on manually-generated messages,
1111 - MAY be used on Delivery Status Notifications (DSNs) and Message
1112 Disposition Notifications (MDNs),
1114 - MUST NOT be used on messages generated by automatic or periodic
1115 processes, except for messages which are automatic responses to
1118 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1119 to the default Received: header definition.
1121 PH/12 Added log selector acl_warn_skipped (default on).
1123 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1124 variables because (a) they are in the wrong storage pool and (b) even if
1125 they were copied, it wouldn't work properly because of the caching.
1127 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1128 checking when decoding. Apparently there are clients that generate
1129 overlong encoded strings. Why am I not surprised?
1131 PH/15 If the first argument of "${if match_address" was not empty, but did not
1132 contain an "@" character, Exim crashed. Now it writes a panic log message
1133 and treats the condition as false.
1135 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1137 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1138 extension "envelope-auth". The code is finished and in agreement with
1139 other implementations, but there is no documentation so far and in fact,
1140 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1141 not changing the active code.
1143 Print executed "if" and "elsif" statements when debugging is used. This
1144 helps a great deal to understand what a filter does.
1146 Document more things not specified clearly in RFC3028. I had all this
1147 sorted out, when out of a sudden new issues came to my mind. Oops."
1149 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1152 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1153 canonical form (as documented). However, after a host name lookup from
1154 the IP address, check_host() was doing a simple string comparison with
1155 addresses acquired from the DNS when checking that the found name did
1156 have the original IP as one of its addresses. Since any found IPv6
1157 addresses are likely to be in abbreviated form, the comparison could
1158 fail. Luckily, there already exists a function for doing the comparison
1159 by converting both addresses to binary, so now that is used instead of
1160 the text comparison.
1162 PH/20 There was another similar case to PH/19, when a complete host name was
1163 given in a host list; looking up its IP address could give an abbreviated
1164 form, whereas the current host's name might or might not be abbreviated.
1165 The same fix has been applied.
1171 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1172 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1175 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1176 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1178 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1180 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1181 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1182 Also, for "Auto-Submitted", treat anything other than "no" as
1183 non-personal, in accordance with RFC 3834. (Previously it treated
1184 anything starting "auto-" as non-personal.)
1186 TF/01 The control=submission/name=... option had a problem with syntax
1187 errors if the name included a slash character. The /name= option
1188 now slurps the rest of the string, so it can include any characters
1189 but it must come last in the list of options (after /sender_retain
1192 PH/05 Some modifications to the interface to the fake nameserver for the new
1200 TK/01 Added the "success_on_redirect" address verification option. See
1201 NewStuff for rationale and an example.
1203 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1205 PH/02 Patch to exigrep to allow it to work on syslog lines.
1207 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1208 fread() to skip over the body file's header line, because in Cygwin the
1209 header line is locked and is inaccessible.
1211 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1212 co-exist for some time) to make it clear that it is the Exim ID that is
1213 referenced, not the Message-ID: header line.
1215 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1216 string_format() function, because snprintf() does not exist on all
1219 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1220 ${stat: expansion item.
1222 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1223 protocol synchronization error", to keep the pedants happy.
1225 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1226 well as for IRIX systems, when gcc is being used. See the host.c source
1229 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1231 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1234 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1235 transport and to the smtp transport in LMTP mode.
1237 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1239 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1240 run a filter in a subprocess. This could lead to confusion in subsequent
1241 lookups in the parent process. There should also be a search_tidyup() at
1242 the end of the subprocess.
1244 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1245 only if the host matched helo_try_verify_hosts, which caused the
1246 verification to occur when the EHLO/HELO command was issued. The ACL just
1247 tested the remembered result. Now, if a previous verification attempt has
1248 not happened, "verify = helo" does it there and then.
1250 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1252 TK/03 Fix log output including CR from clamd.
1254 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1255 exist provoked a memory error which could cause a segfault.
1257 PH/15 Installed PCRE 6.2
1259 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1261 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1262 of the problem. Specifically, suggested +O2 rather than +O1 for the
1265 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1267 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1268 record, the port was not correctly logged when the outgoing_port log
1269 selector was set (it logged the transort's default port).
1271 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1272 fallback_hosts, and "hosts" in the smtp transport.
1274 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1275 host errors such as "Connection refused".
1277 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1278 authentication with radiusclient 0.4.9:
1280 - Error returned from rc_read_config was caught wrongly
1281 - Username/password not passed on to radius server due to wrong length.
1283 The presumption is that some radiusclient API changes for 4.51/PH/17
1284 were not taken care of correctly. The code is still untested by me (my
1285 Linux distribution still has 0.3.2 of radiusclient), but it was
1286 contributed by a Radius user.
1288 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1289 expanding the "port" option of the smtp transport.
1291 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1292 while reading a MIME header. Thanks to Tom Hughes for a patch.
1294 PH/24 Include config.h inside local_scan.h so that configuration settings are
1297 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1298 that in an ACL for RCPT (for example), you can examine exactly what was
1301 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1302 commands, but it was not correctly comparing the address with the actual
1303 client host address. Thus, it would show the EHLO address in Received:
1304 header lines when this was not necessary.
1306 PH/27 Added the % operator to ${eval:}.
1308 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1309 it should be ignoring failures (because with -C, for example, it has lost
1310 privilege). It wasn't ignoring creation failures other than "already
1313 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1316 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1317 because an input line was too long, either on its own, or by virtue of
1318 too many continuations, the temporary file was not being removed, and the
1319 return code was incorrect.
1321 PH/31 Missing "BOOL" in function definition in filtertest.c.
1323 PH/32 Applied Sieve patches from the maintainer.
1325 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1327 PH/33 Added "verify = not_blind".
1329 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1330 Local/Makefile (with some defaults set). These are used in built scripts
1331 such as exicyclog, but they have never been used in the exim_install
1332 script (though there are many overriding facilities there). I have
1333 arranged that the exim_install script now takes note of these two
1336 PH/35 Installed configuration files for Dragonfly.
1338 PH/36 When a locally submitted message by a trusted user did not contain a
1339 From: header, and the sender address was obtained from -f or from an SMTP
1340 MAIL command, and the trusted user did not use -F to supply a sender
1341 name, $originator_name was incorrectly used when constructing a From:
1342 header. Furthermore, $originator_name was used for submission mode
1343 messages from external hosts without From: headers in a similar way,
1344 which is clearly wrong.
1346 PH/37 Added control=suppress_local_fixups.
1348 PH/38 When log_selector = +received_sender was set, and the addition of the
1349 sender made the log line's construction buffer exactly full, or one byte
1350 less than full, an overflow happened when the terminating "\n" was
1353 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1354 when the result of a list match is failure because a DNS lookup failed.
1356 PH/40 RM_COMMAND is now used in the building process.
1358 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1359 the "build-* directories that it finds.
1361 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1362 address in a domain literal was a prefix of an interface address.
1364 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1365 when verifying a sender address, unless rewrite_headers is false.
1367 PH/44 Wrote a long comment about why errors_to addresses are verified as
1368 recipients, not senders.
1370 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1371 the ratelimit ACL was added.
1373 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1375 PH/46 Added extra information about PostgreSQL errors to the error string.
1377 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1378 avoiding the need to install special zones in a real server. This is
1379 backwards compatible; if it can't find the fake resolver, it drops back.
1380 Thus, both old and new test suites can be run.
1382 TF/02 Added util/ratelimit.pl
1384 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1385 clock is set back in time.
1387 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1388 Candler <B.Candler@pobox.com>.
1390 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1391 OK for addresses that are the result of redirections.
1393 PH/48 A number of further additions for the benefit of the new test suite,
1394 including a fake gethostbyname() that interfaces to the fake DNS resolver
1397 TF/06 The fix for widen_domains has also been applied to qualify_single and
1398 search_parents which are the other dnslookup options that can cause
1401 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1404 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1406 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1407 service" if it encountered an MX record with an empty target hostname.
1408 The message is now "an MX or SRV record indicated no SMTP service".
1410 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1411 if the DNS of the sending site is misconfigured. This is quite a
1412 common situation. This change restores the behaviour of treating a
1413 helo verification defer as a failure.
1415 PH/51 If self=fail was set on a router, the bounce message did not include the
1416 actual error message.
1422 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1424 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1425 command itself ended in error, the underlying message about the transport
1426 filter timeout was being overwritten with the pipe command error. Now the
1427 underlying error message should be appended to the second error message.
1429 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1431 PH/02 Reduce the amount of output that "make" produces by default. Full output
1432 can still be requested.
1434 PH/03 The warning log line about a condition test deferring for a "warn" verb
1435 was being output only once per connection, rather than after each
1436 occurrence (because it was using the same function as for successful
1437 "warn" verbs). This seems wrong, so I have changed it.
1439 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1440 it should not have, which might have caused a crash in the right
1441 circumstances, but probably never did.
1443 PH/04 Installed a modified version of Tony Finch's patch to make submission
1444 mode fix the return path as well as the Sender: header line, and to
1445 add a /name= option so that you can make the user's friendly name appear
1448 TF/03 Added the control = fakedefer ACL modifier.
1450 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1451 Mark Lowes for thorough testing.
1453 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1455 TK/03 Merged latest SRS patch from Miles Wilton.
1457 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1458 with the definition in sysexits.h (which is #included earlier).
1459 Fortunately, Exim does not actually use EX_OK. The code used to try to
1460 preserve the sysexits.h value, by assumimg that macro definitions were
1461 scanned for macro replacements. I have been disabused of this notion,
1462 so now the code just undefines EX_OK before #including unistd.h.
1464 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1465 in the smtp transport. When a block could not be written in a single
1466 write() function, the timeout was being re-applied to each part-write.
1467 This seems wrong - if the receiver was accepting one byte at a time it
1468 would take for ever. The timeout is now adjusted when this happens. It
1469 doesn't have to be particularly precise.
1471 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1472 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1474 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1475 without a domain if the check to <postmaster@domain> fails.
1477 SC/01 Eximstats: added -xls and the ability to specify output files
1478 (patch written by Frank Heydlauf).
1480 SC/02 Eximstats: use FileHandles for outputing results.
1482 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1484 SC/04 Eximstats: fixed display of large numbers with -nvr option
1486 SC/05 Eximstats: fixed merging of reports with empty tables.
1488 SC/06 Eximstats: added the -include_original_destination flag
1490 SC/07 Eximstats: removed tabs and trailing whitespace.
1492 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1494 TK/06 MBOX spool code: Add real "From " MBOX separator line
1495 so the .eml file is really in mbox format (even though
1496 most programs do not really care). Patch from Alex Miller.
1498 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1499 The latter is generated from $received_to and is only set if the
1500 message has one envelope recipient. SA can use these headers,
1501 obviously out-of-the-box. Patch from Alex Miller.
1503 PH/08 The ${def test on a variable was returning false if the variable's
1504 value was "0", contrary to what the specification has always said!
1505 The result should be true unless the variable is empty.
1507 PH/09 The syntax error of a character other than { following "${if
1508 def:variable_name" (after optional whitespace) was not being diagnosed.
1509 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1510 accidental colon was present, for example, could give incorrect results.
1512 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1513 result is used (not including appendfile, where other changes are about
1516 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1517 This involved changing a lot of size variables from int to off_t. It
1518 should work with maildirs and everything.
1520 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1521 spamd dying while we are connected to it.
1523 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1526 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1527 function for BDB 4.3.
1529 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1531 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1532 the sources for PCRE have been re-arranged and more files are now
1535 PH/15 The code I had for printing potentially long long variables in PH/11
1536 above was not the best (it lost precision). The length of off_t variables
1537 is now inspected at build time, and an appropriate printing format (%ld
1538 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1539 to be "long long int" or "long int". This is needed for the internal
1540 formatting function string_vformat().
1542 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1543 the configuration file to be ":syslog", then the script "guesses" where
1544 the logs files are, rather than using the compiled in default. In our
1545 case the guess is not the same as the compiled default, so the script
1546 suddenly stopped working when I started to use syslog. The patch checks
1547 to see if log_file_path is "". If so, it attempts to read it from exim
1548 with no configuration file to get the compiled in version, before it
1549 falls back to the previous guessing code."
1551 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1552 implementing BATV in an Exim configuration. See NewStuff for the gory
1555 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1556 Makefile that are specific to HP-UX.
1558 PH/18 If the "use_postmaster" option was set for a recipient callout together
1559 with the "random" option, the postmaster address was used as the MAIL
1560 FROM address for the random test, but not for the subsequent recipient
1561 test. It is now used for both.
1563 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1564 patch removes a few documentation additions to RFC 3028, because the
1565 latest draft now contains them. It adds the new en;ascii-case comparator
1566 and a new error check for 8bit text in MIME parts. Comparator and
1567 require names are now matched exactly. I enabled the subaddress
1568 extension, but it is not well tested yet (read: it works for me)."
1570 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1571 rework some of the code of TK/09 above to avoid the hardwired use of
1572 "%lld" and "long long". Replaced the call to snprintf() with a call to
1575 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1576 records point to non-existent hosts", "retry timeout exceeded", and
1577 "retry time not reached for any host after a long failure period".
1579 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1580 experimental DomainKeys support:
1582 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1583 (2) On an error such as an illegally used "control", the wrong name for
1584 the control was given.
1586 These problems did NOT occur unless DomainKeys support was compiled.
1588 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1590 PH/24 Added ${if match_ip condition.
1592 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1593 databases so that it will be absolutely obvious if a crash occurs in the
1594 DB library. This is a regular occurrence (often caused by mis-matched
1597 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1598 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1599 picked up on a user's system that detects such things. There doesn't seem
1600 to be a gcc warning option for this - only an attribute that has to be
1601 put on the function's prototype. It seems that in Fedora Core 4 they have
1602 set this on a number of new functions. No doubt there will be more in due
1605 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1606 specify a transport. However, if an address that was verified by such a
1607 router was the subject of a callout, Exim crashed because it tried to
1608 read the rcpt_include_affixes from the non-existent transport. Now it
1609 just assumes that the setting of that option is false. This bug was
1610 introduced by 4.51/PH/31.
1612 PH/28 Changed -d+all to exclude +memory, because that information is very
1613 rarely of interest, but it makes the output a lot bigger. People tend to
1614 do -d+all out of habit.
1616 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1617 code in os-type was giving problems when libc.so lives in lib64, like on
1620 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1621 aren't the modern standard, and it seems that some systems' include files
1622 don't always have them. Exim was already checking for some of the newer
1623 ones like T_AAAA, and defining it itself. I've added checks for all the
1624 record types that Exim uses.
1626 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1627 not automatically generating a new one, as it is supposed to. This
1628 prevented TLS from working. If the file did exist, but contained invalid
1629 data, a new version was generated, as expected. It was only the case of a
1630 non-existent file that was broken.
1632 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1633 with a change in libdomainkeys > 0.64.
1635 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1636 from DNS. If the selector record carries the flag, it now has
1637 precedence over the domain-wide flag.
1639 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1641 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1642 the use of a port name that isn't defined in /etc/services) occurred, the
1643 message was deferred as in a normal delivery, and thus remained on the
1644 spool, instead of being failed because of the mua_wrapper setting. This
1645 is now fixed, and I tidied up some of the mua_wrapper messages at the
1648 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1649 the messages in a hash of arrays rather than using individual hashes.
1650 This is a bit cleaner and results in dramatic memory savings, albeit
1651 at a slight CPU cost.
1653 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1654 as requested by Marc Sherman.
1656 SC/10 Eximstats: added histograms for user specified patterns as requested
1659 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1661 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1662 fopen() in the content-scanning modules that did not already have it.
1668 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1669 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1671 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1673 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1675 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1676 to test Sieve filters that use "vacation".
1678 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1679 that changes the way the GnuTLS parameters are stored in the cache file.
1680 The new format can be generated externally. For backward compatibility,
1681 if the data in the cache doesn't make sense, Exim assumes it has read an
1682 old-format file, and it generates new data and writes a new file. This
1683 means that you can't go back to an older release without removing the
1686 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1687 work if there are any delivery delays because "one_time" forces the
1688 parent to be marked "delivered", so its unseen clone is never tried
1689 again. For this reason, Exim now forbids the simultaneous setting of
1692 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1693 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1694 entry if you want to know the details. The fix, however, neglected to
1695 consider the case when local delivery batching is involved. The test for
1696 "previously delivered" was not happening when checking to see if an
1697 address could be batched with a previous (undelivered) one; under
1698 certain circumstances this could lead to multiple deliveries to the same
1701 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1702 in its include files, and this causes problems building Exim.
1704 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1705 header_syntax) but Exim was just ignoring anything given after a slash.
1706 In particular, this caused confusion with an attempt to use "verify =
1707 reverse_host_lookup/defer_ok". An error is now given when options are
1708 supplied for verify items that do not have them. (Maybe reverse_host_
1709 lookup should have a defer_ok option, but that's a different point.)
1711 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1712 defined by RFC 821) to 2048, because there were problems with some AUTH
1713 commands, and RFC 1869 says the size should be increased for extended
1714 SMTP commands that take arguments.
1716 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1719 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1720 "unknown" error; now it says that the functionality isn't in the binary.
1722 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1723 an address' error message when a string expansion fails (syntax or
1724 whatever). Otherwise the password may appear in the log. Following change
1725 PH/42 below, there is no longer a chance of it appearing in a bounce
1728 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1730 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1731 of its addresses was ever tried. (Bugzilla bug #2.)
1733 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1734 the result incorrectly in the debug output. (It correctly added a newline
1735 to what was transported.)
1737 TF/01 Added $received_time.
1739 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1740 commented out examples of how to interface to a virus scanner and to
1741 SpamAssassin. Also added commented examples of av_scanner and
1742 spamd_address settings.
1744 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1745 and controls are allowed in which ACLs. There were a couple of minor
1746 errors. Some of the entries in the conditions table (which is a table of
1747 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1748 negation of where the condition IS allowed.
1750 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1752 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1753 header file does not have a version number, so I've had to invent a new
1754 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1755 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1756 radiusclient), but it was contributed by a Radius user.
1758 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1759 files or directories, for OpenSSL.
1761 PH/19 When an Exim process that is running as root has to create an Exim log
1762 file, it does so in a subprocess that runs as exim:exim so as to get the
1763 ownership right at creation (otherwise, other Exim processes might see
1764 the file with the wrong ownership). There was no test for failure of this
1765 fork() call, which would lead to the process getting stuck as it waited
1766 for a non-existent subprocess. Forks do occasionally fail when resources
1767 run out. I reviewed all the other calls to fork(); they all seem to check
1770 PH/20 When checking for unexpected SMTP input at connect time (before writing
1771 the banner), Exim was not dealing correctly with a non-positive return
1772 from the read() function. If the client had disconnected by this time,
1773 the result was a log entry for a synchronization error with an empty
1774 string after "input=" when read() returned zero. If read() returned -1
1775 (an event I could not check), uninitialized data bytes were printed.
1776 There were reports of junk text (parts of files, etc) appearing after
1779 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1781 PH/22 Added support for macro redefinition, and (re)definition in between
1782 driver and ACL definitions.
1784 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1785 forgetting to use the resulting value; it was using the unexpanded value.
1787 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1788 hadn't been configured. The fix is from Juergen Kreileder, who
1789 understands it better than I do:
1791 "Here's what I see happening with three configured cyrus_sasl
1792 authenticators configured (plain, login, cram-md5):
1794 On startup auth_cyrus_sasl_init() gets called for each of these.
1795 This means three calls to sasl_listmech() without a specified mech_list.
1796 => SASL tests which mechs of all available mechs actually work
1797 => three warnings about OTP not working
1798 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1800 With the patch, sasl_listmech() also gets called three times. But now
1801 SASL's mech_list option is set to the server_mech specified in the the
1802 authenticator. Or in other words, the answer from sasl_listmech()
1803 gets limited to just the mech you're testing for (which is different
1805 => the return list contains just 'plain' or 'login', 'cram-md5' or
1806 nothing depending on the value of ob->server_mech.
1808 I've just tested the patch: Authentication still works fine,
1809 unavailable mechs specified in the exim configuration are still
1810 caught, and the auth.log warnings about OTP are gone."
1812 PH/25 When debugging is enabled, the contents of the command line are added
1813 to the debugging output, even when log_selector=+arguments is not
1816 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1817 answer is "GNU", and only if the return is "GNU/something" is the answer
1820 PH/27 $acl_verify_message is now set immediately after the failure of a
1821 verification in an ACL, and so is available in subsequent modifiers. In
1822 particular, the message can be preserved by coding like this:
1824 warn !verify = sender
1825 set acl_m0 = $acl_verify_message
1827 Previously, $acl_verify_message was set only while expanding "message"
1828 and "log_message" when a very denied access.
1830 PH/28 Modified OS/os.c-Linux with
1832 -#ifndef OS_LOAD_AVERAGE
1833 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1835 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1836 nomenclature these days.)
1838 PH/29 Installed patch from the Sieve maintainer that adds the options
1839 sieve_useraddress and sieve_subaddress to the redirect router.
1841 PH/30 In these circumstances:
1842 . Two addresses routed to the same list of hosts;
1843 . First host does not offer TLS;
1844 . First host accepts first address;
1845 . First host gives temporary error to second address;
1846 . Second host offers TLS and a TLS session is established;
1847 . Second host accepts second address.
1848 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1849 and peerdn, if requested) that were in fact used only for the second
1852 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1853 attention to any local part prefix or suffix that was matched by the
1854 router that accepted the address. It now behaves in the same way as it
1855 does for delivery: the affixes are removed from the local part unless
1856 rcpt_include_affixes is set on the transport.
1858 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1859 timeout during the DATA phase of an incoming message.
1861 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1862 applied a patch sanctioned by the Sieve maintainer.
1864 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1865 the uid or gid is negative. A case of a negative gid caused this to be
1866 noticed. The fix allows for either to be negative.
1868 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1869 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1872 PH/36 The change PH/12 above was broken. Fixed it.
1874 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1875 the grounds that routing the same address twice would always produce the
1876 same answer. This might have been true once, but it is certainly no
1877 longer true now. Routing a child address may depend on the previous
1878 routing that produced that child. Some complicated redirection strategies
1879 went wrong when messages had multiple recipients, and made Exim's
1880 behaviour dependent on the order in which the addresses were given.
1882 I have moved the duplicate checking until after the routing is complete.
1883 Exim scans the addresses that are assigned to local and remote
1884 transports, and removes any duplicates. This means that more work will be
1885 done, as duplicates will always all be routed, but duplicates are
1886 presumably rare, so I don't expect this is of any significance.
1888 For deliveries to pipes, files, and autoreplies, the duplicate checking
1889 still happens during the routing process, since they are not going to be
1892 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1893 It corrects a timeout issue with spamd. This is Ian's comment: "The
1894 background is that sometimes spamd either never reads data from a
1895 connection it has accepted, or it never writes response data. The exiscan
1896 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1897 blindly assumes that writes won't block so it may never time out."
1899 PH/39 Allow G after quota size as well as K and M.
1901 PH/40 The value set for $authenticated_id in an authenticator may not contain
1902 binary zeroes or newlines because the value is written to log lines and
1903 to spool files. There was no check on this. Now the value is run through
1904 the string_printing() function so that such characters are converted to
1905 printable escape sequences.
1907 PH/41 $message_linecount is a new variable that contains the total number of
1908 lines in the message. Compare $body_linecount, which is the count for the
1911 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1912 bounce and delay warning messages, except in certain special cases, which
1915 (a) An SMTP error message from a remote host;
1916 (b) A message specified in a :fail: redirection;
1917 (c) A message specified in a "fail" command in a system filter;
1918 (d) A message specified in a FAIL return from the queryprogram router;
1919 (e) A message specified by the cannot_route_message router option.
1921 In these cases only, Exim does include the error details in bounce and
1922 warning messages. There are also a few cases where bland messages such
1923 as "unrouteable address" or "local delivery error" are given.
1925 PH/43 $value is now also set for the "else" part of a ${run expansion.
1927 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1928 being worked on, but at least Exim now implements the latest version to
1931 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1932 process to complete was treated as a delivery failure, a timeout while
1933 writing the message to the pipe was logged, but erroneously treated as a
1934 successful delivery. Such timeouts include transport filter timeouts. For
1935 consistency with the overall process timeout, these timeouts are now
1936 treated as errors, giving rise to delivery failures by default. However,
1937 there is now a new Boolean option for the pipe transport called
1938 timeout_defer, which, if set TRUE, converts the failures into defers for
1939 both kinds of timeout. A transport filter timeout is now identified in
1942 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1943 systems where "make" and "gmake" are different, calling "gmake" at top
1944 level broke things. I've arranged for the value of $(MAKE) to be passed
1945 from the Makefile to this script so that it can call the same version of
1949 A note about Exim versions 4.44 and 4.50
1950 ----------------------------------------
1952 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1953 changes of various kinds. As a consequence, a big documentation update was
1954 needed. This delayed the release for rather longer than seemed good, especially
1955 in the light of a couple of (minor) security issues. Therefore, the changes
1956 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1957 release. So 4.44 and 4.50 are in effect two different branches that both start
1960 I have left the 4.50 change log unchanged; it contains all the changes since
1961 4.43. The change log for 4.44 is below; many of its items are identical to
1962 those for 4.50. This seems to be the most sensible way to preserve the
1963 historical information.
1969 1. Minor wording change to the doc/README.SIEVE file.
1971 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1972 computation of the current number of files was incorrect.
1974 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1975 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1978 4. Give more explanation in the error message when the command for a transport
1979 filter fails to execute.
1981 5. There are several places where Exim runs a non-Exim command in a
1982 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1983 was being done only for the command run by the queryprogram router. It is
1984 now done for all such subprocesses. The other cases are: ${run, transport
1985 filters, and the commands run by the lmtp and pipe transports.
1987 6. Added CONFIGURE_GROUP build-time option.
1989 7. Some older OS have a limit of 256 on the maximum number of file
1990 descriptors. Exim was using setrlimit() to set 1000 as a large value
1991 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1992 systems. I've change it so that if it can't get 1000, it tries for 256.
1994 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1995 was an oversight, and furthermore, ever since the addition of extra
1996 controls (e.g. 4.43/32), the checks on when to allow different forms of
1997 "control" were broken. There should now be diagnostics for all cases when a
1998 control that does not make sense is encountered.
2000 9. Added the /retain_sender option to "control=submission".
2002 10. $recipients is now available in the predata ACL (oversight).
2004 11. Tidy the search cache before the fork to do a delivery from a message
2005 received from the command line. Otherwise the child will trigger a lookup
2006 failure and thereby defer the delivery if it tries to use (for example) a
2007 cached ldap connection that the parent has called unbind on.
2009 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2010 of $address_data from the recipient verification was clobbered by the
2011 sender verification.
2013 13. The value of address_data from a sender verification is now available in
2014 $sender_address_data in subsequent conditions in the ACL statement.
2016 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2018 15. Added a new option "connect=<time>" to callout options, to set a different
2021 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2022 was its contents. (It was OK if the option was not defined at all.)
2024 17. A "Completed" log line is now written for messages that are removed from
2025 the spool by the -Mrm option.
2027 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2028 information about exactly what failed.
2030 19. Added -dd to debug only the daemon process.
2032 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2033 handles timeouts, both on the server side and network timeouts. Renamed the
2034 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2036 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2037 It is now set to "smtps".
2039 22. $host_address is now set to the target address during the checking of
2040 ignore_target_hosts.
2042 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2043 being passed; this would have caused $sender_host_name to have been used if
2044 matching the list had actually called for a host name (not very likely,
2045 since this list is usually IP addresses). A host name is now passed as
2048 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2049 code that specifies a non-restarting handler (typically sigaction() in
2050 modern systems) in an attempt to fix a rare and obscure crash bug.
2052 25. Narrowed the window for a race in the daemon that could cause it to ignore
2053 SIGCHLD signals. This is not a major problem, because they are used only to
2054 wake it up if nothing else does.
2056 26. A malformed maildirsize file could cause Exim to calculate negative values
2057 for the mailbox size or file count. Odd effects could occur as a result.
2058 The maildirsize information is now recalculated if the size or filecount
2061 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2062 support for a long time. Removed HAVE_SYS_VFS_H.
2064 28. Installed the latest version of exipick from John Jetmore.
2066 29. In an address list, if the pattern was not a regular expression, an empty
2067 subject address (from a bounce message) matched only if the pattern was an
2068 empty string. Non-empty patterns were not even tested. This was the wrong
2069 because it is perfectly reasonable to use an empty address as part of a
2070 database query. An empty address is now tested by patterns that are
2071 lookups. However, all the other forms of pattern expect the subject to
2072 contain a local part and a domain, and therefore, for them, an empty
2073 address still always fails if the pattern is not itself empty.
2075 30. Exim went into a mad DNS loop when attempting to do a callout where the
2076 host was specified on an smtp transport, and looking it up yielded more
2077 than one IP address.
2079 31. Re-factored the code for checking spool and log partition space into a
2080 function that finds that data and another that does the check. The former
2081 is then used to implement four new variables: $spool_space, $log_space,
2082 $spool_inodes, and $log_inodes.
2084 32. The RFC2047 encoding function was originally intended for short strings
2085 such as real names; it was not keeping to the 75-character limit for
2086 encoded words that the RFC imposes. It now respects the limit, and
2087 generates multiple encoded words if necessary. To be on the safe side, I
2088 have increased the buffer size for the ${rfc2047: expansion operator from
2091 33. It is now permitted to omit both strings after an "if" condition; if the
2092 condition is true, the result is "true". As before, when the second string
2093 is omitted, a false condition yields an empty string. This makes it less
2094 cumbersome to write custom ACL and router conditions.
2096 34. Failure to deliver a bounce message always caused it to be frozen, even if
2097 there was an errors_to setting on the router. The errors_to setting is now
2100 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2101 canonical form (fully expanded) before being placed in
2102 $sender_host_address.
2104 36. The table in the code that translates DNS record types into text (T_A to
2105 "A" for instance) was missing entries for NS and CNAME. It is just possible
2106 that this could have caused confusion if both these types were looked up
2107 for the same domain, because the text type is used as part of Exim's
2108 per-process caching. But the chance of anyone hitting this buglet seems
2111 37. The dnsdb lookup has been extended in a number of ways.
2113 (1) There is a new type, "zns", which walks up the domain tree until it
2114 finds some nameserver records. It should be used with care.
2116 (2) There is a new type, "mxh", which is like "mx" except that it returns
2117 just the host names, not the priorities.
2119 (3) It is now possible to give a list of domains (or IP addresses) to be
2120 looked up. The behaviour when one of the lookups defers can be
2121 controlled by a keyword.
2123 (4) It is now possible to specify the separator character for use when
2124 multiple records are returned.
2126 38. The dnslists ACL condition has been extended: it is now possible to supply
2127 a list of IP addresses and/or domains to be looked up in a particular DNS
2130 39. Added log_selector=+queue_time_overall.
2132 40. When running the queue in the test harness, wait just a tad after forking a
2133 delivery process, to get repeatability of debugging output.
2135 41. Include certificate and key file names in error message when GnuTLS fails
2136 to set them up, because the GnuTLS error message doesn't include the name
2137 of the failing file when there is a problem reading it.
2139 42. Allow both -bf and -bF in the same test run.
2141 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2143 44. The "Exiscan patch" is now merged into the mainline Exim source.
2145 45. Sometimes the final signoff response after QUIT could fail to get
2146 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2147 < 0 before doing a fflush(). This bug looks as though it goes back to the
2148 introduction of TLS in release 3.20, but "sometimes" must have been rare
2149 because the tests only now provoked it.
2151 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2152 (this can affect the format of dates).
2154 47. exim_tidydb, when checking for the continued existence of a message for
2155 which it has found a message-specific retry record, was not finding
2156 messages that were in split spool directories. Consequently, it was
2157 deleting retry records that should have stayed in existence.
2159 48. Steve fixed some bugs in eximstats.
2161 49. The SPA authentication driver was not abandoning authentication and moving
2162 on to the next authenticator when an expansion was forced to fail,
2163 contradicting the general specification for all authenticators. Instead it
2164 was generating a temporary error. It now behaves as specified.
2166 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2167 (the order specifies the preference for clients). The order is now AES256,
2168 AES128, 3DES, ARCFOUR128.
2170 51. Small patch to Sieve code - explicitly set From: when generating an
2173 52. Exim crashed if a remote delivery caused a very long error message to be
2174 recorded - for instance if somebody sent an entire SpamAssassin report back
2175 as a large number of 550 error lines. This bug was coincidentally fixed by
2176 increasing the size of one of Exim's internal buffers (big_buffer) that
2177 happened as part of the Exiscan merge. However, to be on the safe side, I
2178 have made the code more robust (and fixed the comments that describe what
2181 53. Now that there can be additional text after "Completed" in log lines (if
2182 the queue_time_overall log selector is set), a one-byte patch to exigrep
2183 was needed to allow it to recognize "Completed" as not the last thing in
2186 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2187 patch that reportedly fixes this has been added. I am not expert enough to
2188 create a test for it. This is what the patch creator wrote:
2190 "I found a little strange behaviour of ldap code when working with
2191 Windows 2003 AD Domain, where users was placed in more than one
2192 Organization Units. When I tried to give exim partial DN, the exit code
2193 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2194 But simultaneously result of request was absolutely normal ldap result,
2195 so I produce this patch..."
2197 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2198 REFERENCE, so I have modified the code to exclude the patch when that macro
2201 55. Some experimental protocols are using DNS PTR records for new purposes. The
2202 keys for these records are domain names, not reversed IP addresses. The
2203 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2204 leaves it alone. Component reversal etc. now happens only for IP addresses.
2207 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2209 57. Double the size of the debug message buffer (to 2048) so that more of very
2210 long debug lines gets shown.
2212 58. The exicyclog utility now does better if the number of log files to keep
2213 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2215 59. Two changes related to the smtp_active_hostname option:
2217 (1) $smtp_active_hostname is now available as a variable.
2218 (2) The default for smtp_banner uses $smtp_active_hostname instead
2219 of $primary_hostname.
2221 60. The host_aton() function is supposed to be passed a string that is known
2222 to be a valid IP address. However, in the case of IPv6 addresses, it was
2223 not checking this. This is a hostage to fortune. Exim now panics and dies
2224 if the condition is not met. A case was found where this could be provoked
2225 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2226 components; fortuitously, this particular loophole had already been fixed
2227 by change 4.50/55 above.
2229 If there are any other similar loopholes, the new check in host_aton()
2230 itself should stop them being exploited. The report I received stated that
2231 data on the command line could provoke the exploit when Exim was running as
2232 exim, but did not say which command line option was involved. All I could
2233 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2234 running as the user.
2237 61. There was a buffer overflow vulnerability in the SPA authentication code
2238 (which came originally from the Samba project). I have added a test to the
2239 spa_base64_to_bits() function which I hope fixes it.
2242 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2243 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2245 63. The daemon start-up calls getloadavg() while still root for those OS that
2246 need the first call to be done as root, but it missed one case: when
2247 deliver_queue_load_max is set with deliver_drop_privilege. This is
2248 necessary for the benefit of the queue runner, because there is no re-exec
2249 when deliver_drop_privilege is set.
2251 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2252 This has been fixed.
2254 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2255 was in use, was not putting the data itself into the right store pool;
2256 consequently, it could be overwritten for a subsequent message in the same
2257 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2260 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2262 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2263 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2264 the function were treating the return as a boolean value, which happened to
2265 work because 0=false and not-0=true, but is not correct code.
2267 68. The host_aton() function was not handling scoped IPv6 addresses (those
2268 with, for example, "%eth0" on the end) correctly.
2270 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2271 negated items (that is, ~something) in unsigned ints. Some compilers
2272 apparently mutter when there is no cast.
2274 70. If an address verification called from an ACL failed, and did not produce a
2275 user-specific message (i.e. there was only a "system" message), nothing was
2276 put in $acl_verify_message. In this situation, it now puts the system
2279 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2280 session; change 4.31/43 added the unwanted input to the log line - except
2281 that it did not do this in the start of session case. It now does.
2283 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2284 This is wrong and can cause the other end to generate a synchronization
2285 error if it is another Exim or anything else that does the synchronization
2286 check. A QUIT command is no longer sent after a timeout.
2288 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2289 during host lookups.
2291 74. The defer_ok option of callout verification was not working if it was used
2292 when verifying addresses in header lines, that is, for this case:
2294 verify = header_sender/callout=defer_ok
2296 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2297 those file descriptors could be used for SMTP connections. If anything
2298 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2299 could be sent to the SMTP client, causing chaos. The daemon now opens
2300 stdin, stdout, and stderr to /dev/null when it puts itself into the
2303 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2304 log by default. The user can override this with suitable Perl magic.
2306 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2307 the log message when discard triggers, was not working for the DATA ACL or
2308 for the non-SMTP ACL.
2310 78. Error message wording change in sieve.c.
2312 79. If smtp_accept_max_per_host was set, the number of connections could be
2313 restricted to fewer than expected, because the daemon was trying to set up
2314 a new connection before checking whether the processes handling previous
2315 connections had finished. The check for completed processes is now done
2316 earlier. On busy systems, this bug wouldn't be noticed because something
2317 else would have woken the daemon, and it would have reaped the completed
2320 80. If a message was submitted locally by a user whose login name contained one
2321 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2322 It caused a spool format error. I have fixed the spool reading code. A
2323 related problem was that the "from" clause in the Received: line became
2324 illegal because of the space(s). It is now covered by ${quote_local_part.
2326 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2329 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2331 83. Patch from Sieve maintainer to fix unterminated string problem in
2332 "vacation" handling.
2334 84. Some minor changes to the Linux configuration files to help with other
2335 OS variants using glibc.
2337 85. One more patch for Sieve to update vacation handling to latest spec.
2340 ----------------------------------------------------
2341 See the note above about the 4.44 and 4.50 releases.
2342 ----------------------------------------------------
2348 1. Change 4.43/35 introduced a bug that caused file counts to be
2349 incorrectly computed when quota_filecount was set in an appendfile
2352 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2353 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2356 3. Give more explanation in the error message when the command for a transport
2357 filter fails to execute.
2359 4. There are several places where Exim runs a non-Exim command in a
2360 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2361 was being done only for the command run by the queryprogram router. It is
2362 now done for all such subprocesses. The other cases are: ${run, transport
2363 filters, and the commands run by the lmtp and pipe transports.
2365 5. Some older OS have a limit of 256 on the maximum number of file
2366 descriptors. Exim was using setrlimit() to set 1000 as a large value
2367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2368 systems. I've change it so that if it can't get 1000, it tries for 256.
2370 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2371 was an oversight, and furthermore, ever since the addition of extra
2372 controls (e.g. 4.43/32), the checks on when to allow different forms of
2373 "control" were broken. There should now be diagnostics for all cases when a
2374 control that does not make sense is encountered.
2376 7. $recipients is now available in the predata ACL (oversight).
2378 8. Tidy the search cache before the fork to do a delivery from a message
2379 received from the command line. Otherwise the child will trigger a lookup
2380 failure and thereby defer the delivery if it tries to use (for example) a
2381 cached ldap connection that the parent has called unbind on.
2383 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2384 of $address_data from the recipient verification was clobbered by the
2385 sender verification.
2387 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2388 was its contents. (It was OK if the option was not defined at all.)
2390 11. A "Completed" log line is now written for messages that are removed from
2391 the spool by the -Mrm option.
2393 12. $host_address is now set to the target address during the checking of
2394 ignore_target_hosts.
2396 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2397 being passed; this would have caused $sender_host_name to have been used if
2398 matching the list had actually called for a host name (not very likely,
2399 since this list is usually IP addresses). A host name is now passed as
2402 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2403 code that specifies a non-restarting handler (typically sigaction() in
2404 modern systems) in an attempt to fix a rare and obscure crash bug.
2406 15. Narrowed the window for a race in the daemon that could cause it to ignore
2407 SIGCHLD signals. This is not a major problem, because they are used only to
2408 wake it up if nothing else does.
2410 16. A malformed maildirsize file could cause Exim to calculate negative values
2411 for the mailbox size or file count. Odd effects could occur as a result.
2412 The maildirsize information is now recalculated if the size or filecount
2415 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2416 support for a long time. Removed HAVE_SYS_VFS_H.
2418 18. Updated exipick to current release from John Jetmore.
2420 19. Allow an empty sender to be matched against a lookup in an address list.
2421 Previously the only cases considered were a regular expression, or an
2424 20. Exim went into a mad DNS lookup loop when doing a callout where the
2425 host was specified on the transport, if the DNS lookup yielded more than
2428 21. The RFC2047 encoding function was originally intended for short strings
2429 such as real names; it was not keeping to the 75-character limit for
2430 encoded words that the RFC imposes. It now respects the limit, and
2431 generates multiple encoded words if necessary. To be on the safe side, I
2432 have increased the buffer size for the ${rfc2047: expansion operator from
2435 22. Failure to deliver a bounce message always caused it to be frozen, even if
2436 there was an errors_to setting on the router. The errors_to setting is now
2439 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2440 canonical form (fully expanded) before being placed in
2441 $sender_host_address.
2443 24. Updated eximstats to version 1.33
2445 25. Include certificate and key file names in error message when GnuTLS fails
2446 to set them up, because the GnuTLS error message doesn't include the name
2447 of the failing file when there is a problem reading it.
2449 26. Expand error message when OpenSSL has problems setting up cert/key files.
2452 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2453 (this can affect the format of dates).
2455 28. exim_tidydb, when checking for the continued existence of a message for
2456 which it has found a message-specific retry record, was not finding
2457 messages that were in split spool directories. Consequently, it was
2458 deleting retry records that should have stayed in existence.
2460 29. eximstats updated to version 1.35
2461 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2462 1.35 - bugfix such that pie charts by volume are generated correctly
2464 30. The SPA authentication driver was not abandoning authentication and moving
2465 on to the next authenticator when an expansion was forced to fail,
2466 contradicting the general specification for all authenticators. Instead it
2467 was generating a temporary error. It now behaves as specified.
2469 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2470 (the order specifies the preference for clients). The order is now AES256,
2471 AES128, 3DES, ARCFOUR128.
2473 31. Small patch to Sieve code - explicitly set From: when generating an
2476 32. Exim crashed if a remote delivery caused a very long error message to be
2477 recorded - for instance if somebody sent an entire SpamAssassin report back
2478 as a large number of 550 error lines. This bug was coincidentally fixed by
2479 increasing the size of one of Exim's internal buffers (big_buffer) that
2480 happened as part of the Exiscan merge. However, to be on the safe side, I
2481 have made the code more robust (and fixed the comments that describe what
2484 33. Some experimental protocols are using DNS PTR records for new purposes. The
2485 keys for these records are domain names, not reversed IP addresses. The
2486 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2487 leaves it alone. Component reversal etc. now happens only for IP addresses.
2490 34. The host_aton() function is supposed to be passed a string that is known
2491 to be a valid IP address. However, in the case of IPv6 addresses, it was
2492 not checking this. This is a hostage to fortune. Exim now panics and dies
2493 if the condition is not met. A case was found where this could be provoked
2494 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2495 components; fortuitously, this particular loophole had already been fixed
2496 by change 4.50/55 or 4.44/33 above.
2498 If there are any other similar loopholes, the new check in host_aton()
2499 itself should stop them being exploited. The report I received stated that
2500 data on the command line could provoke the exploit when Exim was running as
2501 exim, but did not say which command line option was involved. All I could
2502 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2503 running as the user.
2506 35. There was a buffer overflow vulnerability in the SPA authentication code
2507 (which came originally from the Samba project). I have added a test to the
2508 spa_base64_to_bits() function which I hope fixes it.
2511 36. The daemon start-up calls getloadavg() while still root for those OS that
2512 need the first call to be done as root, but it missed one case: when
2513 deliver_queue_load_max is set with deliver_drop_privilege. This is
2514 necessary for the benefit of the queue runner, because there is no re-exec
2515 when deliver_drop_privilege is set.
2517 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2518 was in use, was not putting the data itself into the right store pool;
2519 consequently, it could be overwritten for a subsequent message in the same
2520 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2523 38. Sometimes the final signoff response after QUIT could fail to get
2524 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2525 < 0 before doing a fflush(). This bug looks as though it goes back to the
2526 introduction of TLS in release 3.20, but "sometimes" must have been rare
2527 because the tests only now provoked it.
2533 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2534 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2535 Now it returns an integer. A number of calls were still expecting a T/F
2536 return. Fortuitously, in all cases, the tests worked in OK situations,
2537 which is the norm. However, things would have gone wrong on any write
2538 failures on the smtp file descriptor. This function is used when sending
2539 messages over SMTP and also when doing verify callouts.
2541 2. When Exim is called to do synchronous delivery of a locally submitted
2542 message (the -odf or -odi options), it no longer closes stderr before doing
2545 3. Implemented the mua_wrapper option.
2547 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2549 5. Implemented the functions header_remove(), header_testname(),
2550 header_add_at_position(), and receive_remove_recipient(), and exported them
2553 6. If an ACL "warn" statement specified the addition of headers, Exim already
2554 inserted X-ACL-Warn: at the start if there was no header name. However, it
2555 was not making this test for the second and subsequent header lines if
2556 there were newlines in the string. This meant that an invalid header could
2557 be inserted if Exim was badly configured.
2559 7. Allow an ACL "warn" statement to add header lines at the start or after all
2560 the Received: headers, as well as at the end.
2562 8. Added the rcpt_4xx retry error code.
2564 9. Added postmaster_mailfrom=xxx to callout verification option.
2566 10. Added mailfrom=xxxx to the callout verification option, for verify=
2569 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2570 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2572 12. Inserted some casts to stop certain compilers warning when using pointer
2573 differences as field lengths or precisions in printf-type calls (mostly
2574 affecting debugging statements).
2576 13. Added optional readline() support for -be (dynamically loaded).
2578 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2579 same clock tick as a message's arrival, so that its received time was the
2580 same as the "first fail" time on the retry record, and that message
2581 remained on the queue past the ultimate address timeout, every queue runner
2582 would try a delivery (because it was past the ultimate address timeout) but
2583 after another failure, the ultimate address timeout, which should have then
2584 bounced the address, did not kick in. This was a "< instead of <=" error;
2585 in most cases the first failure would have been in the next clock tick
2586 after the received time, and all would be well.
2588 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2589 being recognized when the domain list was tested by the match_domain
2590 condition in an expansion string.
2592 16. Added the ${str2b64: operator.
2594 17. Exim was always calling setrlimit() to set a large limit for the number of
2595 processes, without checking whether the existing limit was already
2596 adequate. (It did check for the limit on file descriptors.) Furthermore,
2597 errors from getrlimit() and setrlimit() were being ignored. Now they are
2598 logged to the main and panic logs, but Exim does carry on, to try to do its
2599 job under whatever limits there are.
2601 18. Imported PCRE 5.0.
2603 19. Trivial typo in log message " temporarily refused connection" (the leading
2606 20. If the log selector return_path_on_delivery was set and an address was
2607 redirected to /dev/null, the delivery process crashed because it assumed
2608 that a return path would always be set for a "successful" delivery. In this
2609 case, the whole delivery is bypassed as an optimization, and therefore no
2612 21. Internal re-arrangement: the function for sending a challenge and reading
2613 a response while authentication was assuming a zero-terminated challenge
2614 string. It's now changed to take a pointer and a length, to allow for
2615 binary data in such strings.
2617 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2619 23. Exim was not respecting finduser_retries when seeking the login of the
2620 uid under which it was called; it was always trying 10 times. (The default
2621 setting of finduser_retries is zero.) Also, it was sleeping after the final
2622 failure, which is pointless.
2624 24. Implemented tls_on_connect_ports.
2626 25. Implemented acl_smtp_predata.
2628 26. If the domain in control=submission is set empty, Exim assumes that the
2629 authenticated id is a complete email address when it generates From: or
2630 Sender: header lines.
2632 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2633 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2634 chown and chgrp in /bin and hostname in /usr/bin.
2636 28. Exim was keeping the "process log" file open after each use, just as it
2637 does for the main log. This opens the possibility of it remaining open for
2638 long periods when the USR1 signal hits a daemon. Occasional processlog
2639 errors were reported, that could have been caused by this. Anyway, it seems
2640 much more sensible not to leave this file open at all, so that is what now
2643 29. The long-running daemon process does not normally write to the log once it
2644 has entered its main loop, and it closes the log before doing so. This is
2645 so that log files can straightforwardly be renamed and moved. However,
2646 there are a couple of unusual error situations where the daemon does write
2647 log entries, and I had neglected to close the log afterwards.
2649 30. The text of an SMTP error response that was received during a remote
2650 delivery was being truncated at 512 bytes. This is too short for some of
2651 the long messages that one sometimes sees. I've increased the limit to
2654 31. It is now possible to make retry rules that apply only when a message has a
2655 specific sender, in particular, an empty sender.
2657 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2658 it possible to be selective about when SMTP synchronization is enforced.
2660 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2662 32. Implemented hosts_connection_nolog.
2664 33. Added an ACL for QUIT.
2666 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2669 35. Added mailbox_size and mailbox_filecount to appendfile.
2671 36. Added control = no_multiline_responses to ACLs.
2673 37. There was a bug in the logic of the code that waits for the clock to tick
2674 in the case where the clock went backwards by a substantial amount such
2675 that the microsecond fraction of "now" was more than the microsecond
2676 fraction of "then" (but the whole seconds number was less).
2678 38. Added support for the libradius Radius client library this is found on
2679 FreeBSD (previously only the radiusclient library was supported).
2685 1. When certain lookups returned multiple values in the form name=value, the
2686 quoting of the values was not always being done properly. Specifically:
2687 (a) If the value started with a double quote, but contained no whitespace,
2689 (b) If the value contained whitespace other than a space character (i.e.
2690 tabs or newlines or carriage returns) it was not quoted.
2691 This fix has been applied to the mysql and pgsql lookups by writing a
2692 separate quoting function and calling it from the lookup code. The fix
2693 should probably also be applied to nisplus, ibase and oracle lookups, but
2694 since I cannot test any of those, I have not disturbed their existing code.
2696 2. A hit in the callout cache for a specific address caused a log line with no
2697 reason for rejecting RCPT. Now it says "Previous (cached) callout
2698 verification failure".
2700 3. There was an off-by-one bug in the queryprogram router. An over-long
2701 return line was truncated at 256 instead of 255 characters, thereby
2702 overflowing its buffer with the terminating zero. As well as fixing this, I
2703 have increased the buffer size to 1024 (and made a note to document this).
2705 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2706 when Exim is waiting for an SMTP response from a remote server, Exim
2707 restarts its select() call on the socket, thereby resetting its timeout.
2708 This is not a problem when such interrupts are rare. Somebody set up a cron
2709 job to run exiwhat every 2 minutes, which is less than the normal select()
2710 timeout (5 or 10 minutes). This meant that the select() timeout never
2711 kicked in because it was always reset. I have fixed this by comparing the
2712 time when an interrupt arrives with the time at the start of the first call
2713 to select(). If more time than the timeout has elapsed, the interrupt is
2714 treated as a timeout.
2716 5. Some internal re-factoring in preparation for the addition of Sieve
2717 extensions (by MH). In particular, the "personal" test is moved to a
2718 separate function, and given an option for scanning Cc: and Bcc: (which is
2719 not set for Exim filters).
2721 6. When Exim created an email address using the login of the caller as the
2722 local part (e.g. when creating a From: or Sender: header line), it was not
2723 quoting the local part when it contained special characters such as @.
2725 7. Installed new OpenBSD configuration files.
2727 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2728 try to make them clearer.
2730 9. Callout options, other than the timeout value, were being ignored when
2731 verifying sender addresses in header lines. For example, when using
2733 verify = header_sender/callout=no_cache
2735 the cache was (incorrectly) being used.
2737 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2738 only the Cygwin environment.
2740 11. When return_path_on_delivery was set as a log selector, if different remote
2741 addresses in the same message used different return paths and parallel
2742 remote delivery occurred, the wrong values would sometimes be logged.
2743 (Whenever a remote delivery process finished, the return path value from
2744 the most recently started remote delivery process was logged.)
2746 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2747 lines when AUTH and/or TLS are in use. This is the "received protocol"
2748 field. Exim used to use "asmtp" for authenticated SMTP, without any
2749 indication (in the protocol name) for TLS use. Now it follows the RFC and
2750 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2751 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2752 names appear in log lines as well as in Received: header lines.
2754 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2755 extensions, and comparison tests, and to fix some bugs.
2757 14. Changes to the "personal" filter test:
2759 (1) The test was buggy in that it was just doing the equivalent of
2760 "contains" tests on header lines. For example, if a user's address was
2761 anne@some.where, the "personal" test would incorrectly be true for
2763 To: susanne@some.where
2765 This test is now done by extracting each address from the header in turn,
2766 and checking the entire address. Other tests that are part of "personal"
2767 are now done using regular expressions (for example, to check local parts
2768 of addresses in From: header lines).
2770 (2) The list of non-personal local parts in From: addresses has been
2771 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2772 taken from the Sieve specification recommendations.
2774 (3) If the message contains any header line starting with "List-" it is
2775 treated as non-personal.
2777 (4) The test for "circular" in the Subject: header line has been removed
2778 because it now seems ill-conceived.
2780 15. Minor typos in src/EDITME comments corrected.
2782 16. Installed latest exipick from John Jetmore.
2784 17. If headers_add on a router specified a text string that was too long for
2785 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2786 of string_sprintf() is now avoided.
2788 18. $message_body_size was not set (it was always zero) when running the DATA
2789 ACL and the local_scan() function.
2791 19. For the "mail" command in an Exim filter, no default was being set for
2792 the once_repeat time, causing a random time value to be used if "once" was
2793 specified. (If the value happened to be <= 0, no repeat happened.) The
2794 default is now 0s, meaning "never repeat". The "vacation" command was OK
2795 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2796 (I found it when inspecting the code).
2798 20. There is now an overall timeout for performing a callout verification. It
2799 defaults to 4 times the callout timeout, which applies to individual SMTP
2800 commands during the callout. The overall timeout applies when there is more
2801 than one host that can be tried. The timeout is checked before trying the
2802 next host. This prevents very long delays if there are a large number of
2803 hosts and all are timing out (e.g. when the network connections are timing
2804 out). The value of the overall timeout can be changed by specifying an
2805 additional sub-option for "callout", called "maxwait". For example:
2807 verify = sender/callout=5s,maxwait=20s
2809 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2810 to the end before writing, but this should make it even safer).
2812 22. Exim was forgetting that it had advertised PIPELINING for the second and
2813 subsequent messages on an SMTP connection. It was also not resetting its
2814 memory on STARTTLS and an internal HELO.
2816 23. When Exim logs an SMTP synchronization error within a session, it now
2817 records whether PIPELINING has been advertised or not.
2819 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2820 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2821 rather than long int.
2823 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2825 26. Added the never_mail option to autoreply.
2831 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2832 crash if the getsockname() call failed; this can happen if a connection is
2833 closed very soon after it is established. The problem was simply in the
2834 order in which certain operations were done, causing Exim to try to write
2835 to the SMTP stream before it had set up the file descriptor. The bug has
2836 been fixed by making things happen in the correct order.
2842 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2843 before the connection was closed, thus losing the rejection response.
2845 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2846 some early Solaris releases, but causes trouble in current releases where
2847 socklen_t is defined.
2849 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2852 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2855 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2856 that releasing the top of it at the end releases what was used for sub-
2857 expansions (unless the block got too big). However, discard this block if
2858 the first thing is a variable or header, so that we can use its block when
2859 it is dynamic (useful for very large $message_headers, for example).
2861 6. Lookups now cache *every* query, not just the most recent. A new, separate
2862 store pool is used for this. It can be recovered when all lookup caches are
2863 flushed. Lookups now release memory at the end of their result strings.
2864 This has involved some general refactoring of the lookup sources.
2866 7. Some code has been added to the store_xxx() functions to reduce the amount
2867 of flapping under certain conditions.
2869 8. log_incoming_interface used to affect only the <= reception log lines. Now
2870 it causes the local interface and port to be added to several more SMTP log
2871 lines, for example "SMTP connection from", and rejection lines.
2873 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2875 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2877 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2878 could be overwritten at the end of the current message (or the start of a
2879 new message if it was set in a HELO ACL). The value is now preserved for
2880 the duration of the SMTP connection.
2882 12. If a transport had a headers_rewrite setting, and a matching header line
2883 contained an unqualified address, that address was qualified, even if it
2884 did not match any rewriting rules. The underlying bug was that the values
2885 of the flags that permit the existence of unqualified sender and recipient
2886 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2887 non-local messages, and by -bnq for local messages) were not being
2888 preserved with the message after it was received.
2890 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2891 "next input=" as part of the text comprising the host identity instead of
2892 the correct text. The code was using the same buffer for two different
2893 strings. However, depending on which order the printing function evaluated
2894 its arguments, the bug did not always show up. Under Linux, for example, my
2895 test suite worked just fine.
2897 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2898 doesn't work with some older versions of Perl. It has been changed to "my",
2899 which in any case is probably the better facility to use.
2901 15. A really picky compiler found some instances of statements for creating
2902 error messages that either had too many or two few arguments for the format
2905 16. The size of the buffer for calls to the DNS resolver has been increased
2906 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2907 for addresses that have a lot of PTR records. This alleviates a problem; it
2908 does not fully solve it.
2910 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2911 buffer now truncates the list and logs the incident, which is the same
2912 action as happens when Exim is looking up a host name and its aliases.
2913 Previously in this situation something unpredictable would happen;
2914 sometimes it was "internal error: store_reset failed".
2916 18. If a server dropped the connection unexpectedly when an Exim client was
2917 using GnuTLS and trying to read a response, the client delivery process
2918 crashed while trying to generate an error log message.
2920 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2921 string, for example:
2923 warn message = H1: something\nH2: something
2925 the text was added as a single header line from Exim's point of view
2926 though it ended up OK in the delivered message. However, searching for the
2927 second and subsequent header lines using $h_h2: did not work. This has been
2928 fixed. Similarly, if a system filter added multiple headers in this way,
2929 the routers could not see them.
2931 20. Expanded the error message when iplsearch is called with an invalid key to
2932 suggest using net-iplsearch in a host list.
2934 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2935 ACLs are no longer actually imposed (and a message to that effect is
2938 22. If a "gecos" field in a passwd entry contained escaped characters, in
2939 particular, if it contained a \" sequence, Exim got it wrong when building
2940 a From: or a Sender: header from that name. A second bug also caused
2941 incorrect handling when an unquoted " was present following a character
2942 that needed quoting.
2944 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2945 was not being matched caselessly.
2947 24. Arranged for all hyphens in the exim.8 source to be escaped with
2950 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2951 properly. Recipient callout cache records were still being keyed to include
2952 the sender, even when use_sender was set false. This led to far more
2953 callouts that were necessary. The sender is no longer included in the key
2954 when use_sender is false.
2956 26. Added "control = submission" modifier to ACLs.
2958 27. Added the ${base62d: operator to decode base 62 numbers.
2960 28. dnsdb lookups can now access SRV records.
2962 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2963 the configuration file.
2965 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2966 (-v) mode. This makes the output for a verbose queue run more intelligible.
2968 31. Added a use_postmaster feature to recipient callouts.
2970 32. Added the $body_zerocount variable, containing the number of binary zero
2971 bytes in the message body.
2973 33. The time of last modification of the "new" subdirectory is now used as the
2974 "mailbox time last read" when there is a quota error for a maildir
2977 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2979 35. Added +ignore_unknown as a special item in host lists.
2981 36. Code for decoding IPv6 addresses in host lists is now included, even if
2982 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2983 address was recognized as an IP address, but was then not correctly decoded
2984 into binary, causing unexpected and incorrect effects when compared with
2991 1. Very minor rewording of debugging text in manualroute to say "list of
2992 hosts" instead of "hostlist".
2994 2. If verify=header_syntax was set, and a header line with an unqualified
2995 address (no domain) and a large number of spaces between the end of the
2996 name and the colon was received, the reception process suffered a buffer
2997 overflow, and (when I tested it) crashed. This was caused by some obsolete
2998 code that should have been removed. The fix is to remove it!
3000 3. When running in the test harness, delay a bit after writing a bounce
3001 message to get a bit more predictability in the log output.
3003 4. Added a call to search_tidyup() just before forking a reception process. In
3004 theory, someone could use a lookup in the expansion of smtp_accept_max_
3005 per_host which, without the tidyup, could leave open a database connection.
3007 5. Added the variables $recipient_data and $sender_data which get set from a
3008 lookup success in an ACL "recipients" or "senders" condition, or a router
3009 "senders" option, similar to $domain_data and $local_part_data.
3011 6. Moved the writing of debug_print from before to after the "senders" test
3014 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3015 problems for message scanning, either using a data ACL, or using
3016 local_scan() because the Received: header was not generated till after they
3017 were called (in order to set the time as the time of reception completion).
3018 I have revised the way this works. The header is now generated after the
3019 body is received, but before the ACL or local_scan() are called. After they
3020 are run, the timestamp in the header is updated.
3026 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3027 before starting a queue runner without re-exec. This happened only when
3028 deliver_drop_privilege was set or when the Exim user was set to root. The
3029 effect of the bug was that timeouts during subsequent deliveries caused
3030 crashes instead of being properly handled. The handler is now left at its
3031 default (and expected) setting.
3033 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3034 message, again when deliver_drop_privilege is set or Exim is run as root.
3035 The bug described in (1) was not present in this case, but the tidying up
3036 of the other signals was missing. I have made the two cases consistent.
3038 3. The ignore_target_hosts setting on a manualroute router was being ignored
3039 for hosts that were looked up using the /MX notation.
3041 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3044 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3045 operated on the sender address. After changing the $sender_address to <>
3046 for the sender address verify, Exim was re-instated it as the original
3047 (before rewriting) address, but remembering that it had rewritten it, so it
3048 wasn't rewriting it again. This bug also had the effect of breaking the
3049 sender address verification caching when the sender address was rewritten.
3051 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3052 This has been changed so that if the ip literal address matches
3053 ignore_target_hosts, the router declines.
3055 7. Added expansion conditions match_domain, match_address, and match_local_
3056 part (NOT match_host).
3058 8. The placeholder for the Received: header didn't have a length field set.
3060 9. Added code to Exim itself and to exim_lock to test for a specific race
3061 condition that could lead to file corruption when using MBX delivery. The
3062 issue is with the lockfile that is created in /tmp. If this file is removed
3063 after a process has opened it but before that process has acquired a lock,
3064 there is the potential for a second process to recreate the file and also
3065 acquire a lock. This could lead to two Exim processes writing to the file
3066 at the same time. The added code performs the same test as UW imapd; it
3067 checks after acquiring the lock that its file descriptor still refers to
3068 the same named file.
3070 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3071 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3074 11. Added the smtp_active_hostname option. If used, this will typically be made
3075 to depend on the incoming interface address. Because $interface_address is
3076 not set up until the daemon has forked a reception process, error responses
3077 that can happen earlier (such as "too many connections") no longer contain
3080 12. If an expansion in a condition on a "warn" statement fails because a lookup
3081 defers, the "warn" statement is abandoned, and the next ACL statement is
3082 processed. Previously this caused the whole ACL to be aborted.
3084 13. Added the iplsearch lookup type.
3086 14. Added ident_timeout as a log selector.
3088 15. Added tls_certificate_verified as a log selector.
3090 16. Added a global option tls_require_ciphers (compare the smtp transport
3091 option of the same name). This controls incoming TLS connections.
3093 17. I finally figured out how to make tls_require_ciphers do a similar thing
3094 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3095 before starting the TLS session.
3097 18. Tabs are now shown as \t in -bP output.
3099 19. If the log selector return_path_on_delivery was set, Exim crashed when
3100 bouncing a message because it had too many Received: header lines.
3102 20. If two routers both had headers_remove settings, and the first one included
3103 a superfluous trailing colon, the final name in the first list and the
3104 first name in the second list were incorrectly joined into one item (with a
3105 colon in the middle).
3111 1. Added -C and -D options to the exinext utility, mainly to make it easier
3112 to include in the automated testing, but these could be helpful when
3113 multiple configurations are in use.
3115 2. The exinext utility was not formatting the output nicely when there was
3116 an alternate port involved in the retry record key, nor when there was a
3117 message id as well (for retries that were specific to a specific message
3118 and a specific host). It was also confused by IPv6 addresses, because of
3119 the additional colons they contain. I have fixed the IPv4 problem, and
3120 patched it up to do a reasonable job for IPv6.
3122 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3123 delivery, the log line now contains "pipelined" if PIPELINING was used.
3125 4. An SMTP transport process used to panic and die if the bind() call to set
3126 an explicit outgoing interface failed. This has been changed; it is now
3127 treated in the same way as a connect() failure.
3129 5. A reference to $sender_host_name in the part of a conditional expansion
3130 that was being skipped was still causing a DNS lookup. This no longer
3133 6. The def: expansion condition was not recognizing references to header lines
3134 that used bh_ and bheader_.
3136 7. Added the _cache feature to named lists.
3138 8. The code for checking quota_filecount in the appendfile transport was
3139 allowing one more file than it should have been.
3141 9. For compatibility with Sendmail, the command line option
3149 and sets the incoming protocol and host name (for trusted callers). The
3150 host name and its colon can be omitted when only the protocol is to be set.
3151 Note the Exim already has two private options, -pd and -ps, that refer to
3152 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3153 "s", but I don't think that's a major issue.
3155 10. A number of refactoring changes to the code, none of which should affect
3158 (a) The number of logging options was getting close to filling up the
3159 32-bit word that was used as a bit map. I have split them into two classes:
3160 those that are passed in the argument to log_write(), and those that are
3161 only ever tested independently outside of that function. These are now in
3162 separate 32-bit words, so there is plenty of room for expansion again.
3163 There is no change in the user interface or the logging behaviour.
3165 (b) When building, for example, log lines, the code previously used a
3166 macro that called string_cat() twice, in order to add two strings. This is
3167 not really sufficiently general. Furthermore, there was one instance where
3168 it was actually wrong because one of the argument was used twice, and in
3169 one call a function was used. (As it happened, calling the function twice
3170 did not affect the overall behaviour.) The macro has been replaced by a
3171 function that can join an arbitrary number of extra strings onto a growing
3174 (c) The code for expansion conditions now uses a table and a binary chop
3175 instead of a serial search (which was left over from when there were very
3176 few conditions). Also, it now recognizes conditions like "pam" even when
3177 the relevant support is not compiled in: a suitably worded error message is
3178 given if an attempt is made to use such a condition.
3180 11. Added ${time_interval:xxxxx}.
3182 12. A bug was causing one of the ddress fields not to be passed back correctly
3183 from remote delivery subprocesses. The field in question was not being
3184 subsequently used, so this caused to problems in practice.
3186 13. Added new log selectors queue_time and deliver_time.
3188 14. Might have fixed a bug in maildirsizefile handling that threw up
3189 "unexpected character" debug warnings, and recalculated the data
3190 unnecessarily. In any case, I expanded the warning message to give more
3193 15. Added the message "Restricted characters in address" to the statements in
3194 the default ACL that block characters like @ and % in local parts.
3196 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3197 Three changes have been made:
3199 (a) There was a serious bug; a negative response to MAIL caused the whole
3200 recipient domain to be cached as invalid, thereby blocking all messages
3201 to all local parts at the same domain, from all senders. This bug has
3202 been fixed. The domain is no longer cached after a negative response to
3203 MAIL if the sender used is not empty.
3205 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3208 (c) A new callout option, "use_sender" has been added for people who want
3209 the modified behaviour.
3215 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3218 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3219 indeed breaks things for older releases.
3221 3. Added additional logging to the case where there is a problem reading data
3222 from a filter that is running in a subprocess using a pipe, in order to
3223 try to track down a specific problem.
3225 4. Testing facility fudge: when running in the test harness and attempting
3226 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3227 getting "No route to host". Convert this to a timeout.
3229 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3232 6. Some OS don't have socklen_t but use size_t instead. This affects the
3233 fifth argument of getsockopt() amongst other things. This is now
3234 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3235 can be set for individual OS. I have set it for SunOS5, OSF1, and
3236 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3237 some earlier ones do not.
3239 7. Change 4.30/15 was not doing the test caselessly.
3241 8. The standard form for an IPv6 address literal was being rejected by address
3242 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3243 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3244 this, as well as the form without the "IPv6" on the front (but only when
3245 address literals are enabled, of course).
3247 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3249 10. Exim crashed if a message with an empty sender address specified by -f
3250 encountered a router with an errors_to setting. This could be provoked only
3251 by a command such as
3255 where an empty string was supplied; "<>" did not hit this bug.
3257 11. Installed PCRE release 4.5.
3259 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3260 remained set. It is now erased.
3262 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3263 times from message ids (which are base 36 rather than the normal 62).
3265 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3266 were being counted as actual protocol errors, and logged if the log
3267 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3268 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3269 and DATA following a set of rejected RCPTs do not count as protocol errors.
3270 In other words, Exim assumes they were pipelined, though this may not
3271 actually be the case. Of course, in all cases the client gets an
3272 appropriate error code.
3274 15. If a lookup fails in an ACL condition, a message about the failure may
3275 be available; it is used if testing the ACL cannot continue, because most
3276 such messages specify what the cause of the deferral is. However, some
3277 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3278 that caused an old message to be retained and used if a later statement
3279 caused a defer, replacing the real cause of the deferral.
3281 16. If an IP address had so many PTR records that the DNS lookup buffer
3282 was not large enough to hold them, Exim could crash while trying to process
3283 the truncated data. It now detects and logs this case.
3285 17. Further to 4.21/58, another change has been made: if (and only if) the
3286 first line of a message (the first header line) ends with CRLF, a bare LF
3287 in a subsequent header line has a space inserted after it, so as not to
3288 terminate the header.
3290 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3291 unnecessarily, used atoi() instead of strtol(), and didn't check the
3292 termination when getting file sizes from file names by regex.
3294 19. Completely re-implemented the support for maildirsize files, in the light
3295 of a number of problems with the previous contributed implementation
3296 (4.30/29). In particular:
3298 . If the quota is zero, the maildirsize file is maintained, but no quota is
3301 . If the maildir directory does not exist, it is created before any attempt
3302 to write a maildirsize file.
3304 . The quota value in the file is just a cache; if the quota is changed in
3305 the transport, the new value overrides.
3307 . A regular expression is available for excluding directories from the
3310 20. The autoreply transport checks the characters in options that define the
3311 message's headers; it allows continued headers, but it was checking with
3312 isspace() after an embedded newline instead of explicitly looking for a
3315 21. If all the "regular" hosts to which an address was routed had passed their
3316 expiry times, and had not reached their retry times, the address was
3317 bounced, even if fallback hosts were defined. Now Exim should go on to try
3320 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3321 equivalent code in the SMTP transport. Some hosts send humungous responses
3322 to HELO/EHLO, more than 1024 it seems.
3324 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3325 gives compiler warnings in some environments. I've now done it "properly",
3328 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3329 (because of problems with the built-in one) was declared to return uschar *
3330 instead of char *, causing compiler failure.
3332 25. Fixed a file descriptor leak when processing alias/forward files.
3334 26. Fixed a minor format string issue in dbfn.c.
3336 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3338 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3339 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3342 29. When a sender address is verified, it is cached, to save repeating the test
3343 when there is more than one recipient in a message. However, when the
3344 verification involves a callout, it is possible for different callout
3345 options to be set for different recipients. It is too complicated to keep
3346 track of this in the cache, so now Exim always runs a verification when a
3347 callout is required, relying on the callout cache for the optimization.
3348 The overhead is duplication of the address routing, but this should not be
3351 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3352 to be verified with callout=postmaster, and the main callout worked but the
3353 postmaster check failed, the verification correctly failed. However, if a
3354 subsequent RCPT command asked for sender verification *without* the
3355 postmaster check, incorrect caching caused this verification also to fail,
3358 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3359 it was not caching the DNS options (qualify_single, search_parents) that
3360 were used when the lookup failed. A subsequent lookup with different
3361 options therefore always gave the same answer, though there were cases
3362 where it should not have. (Example: a "domains = !$mx_any" option on a
3363 dnslookup router: the "domains" option is always processed without any
3364 widening, but the router might have qualify_single set.) Now Exim uses the
3365 cached value only when the same options are set.
3367 32. Added John Jetmore's "exipick" utility to the distribution.
3369 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3370 than a timeout (e.g. a certificate is required, and is not provided), an
3371 Exim server now closes the connection immediately. Previously it waited for
3372 the client to close - but if the client is SSL, it seems that they each
3373 wait for each other, leading to a delay before one of them times out.
3375 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3376 maintained 0.8.x compatibility because I don't think many are using it, and
3377 it is clearly obsolete.
3379 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3382 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3383 client certificate was expired. A simple patch fixes this, though I don't
3384 understand the full logic of why the verify callback is called multiple
3387 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3388 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3389 which causes problems with some clients (such as the Certicom SSL Plus
3390 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3391 disables the coutermeasure allowing Eudora to connect."
3393 38. Exim was not checking that a write() to a log file succeeded. This could
3394 lead to Bad Things if a log got too big, in particular if it hit a file
3395 size limit. Exim now panics and dies if it cannot write to a log file, just
3396 as it does if it cannot open a log file.
3398 39. Modified OS/Makefile-Linux so that it now contains
3400 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3402 The two -D definitions ensure that Exim is compiled with large file
3403 support, which makes it possible to handle log files that are bigger than
3406 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3407 instance) a domain was checked against a named list that involved a lookup,
3408 causing $domain_data to be set, then another domain was checked against the
3409 same list, then the first domain was re-checked, the value of $domain_data
3410 after the final check could be wrong. In particular, if the second check
3411 failed, it could be set empty. This bug probably also applied to
3414 41. The strip_trailing_dot option was not being applied to the address given
3415 with the -f command-line option.
3417 42. The code for reading a message's header from the spool was incrementing
3418 $received_count, but never initializing it. This meant that the value was
3419 incorrect (doubled) while delivering a message in the same process in which
3420 it was received. In the most common configuration of Exim, this never
3421 happens - a fresh exec is done - but it can happen when
3422 deliver_drop_privilege is set.
3424 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3425 it now includes up to 150 characters of the unexpected data in the log
3428 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3429 and building data strings. The size of both of these buffers was 10 000
3430 bytes - far larger than anybody would *ever* want, thought I. Needless to
3431 say, somebody hit the limit. I have increased the maximum line length to
3432 20 000 and the maximum data length of concatenated lines to 100 000. I have
3433 also fixed two bugs, because there was no checking on these buffers. Tsk,
3434 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3435 buffer is too small.
3437 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3438 lsearch lookups. Now it does.
3440 46. When parsing a route_list item in a manualroute router, a fixed-length
3441 buffer was used for the list of hosts. I made this 1024 bytes long,
3442 thinking that nobody would ever have a list of hosts that long. Wrong.
3443 Somebody had a whole pile of complicated expansion conditions, and the
3444 string was silently truncated, leading to an expansion error. It turns out
3445 that it is easier to change to an unlimited length (owing to other changes
3446 that have happened since this code was originally written) than to build
3447 structure for giving a limitation error. The length of the item that
3448 expands into the list of hosts is now unlimited.
3450 47. The lsearch lookup could not handle data where the length of text line was
3451 more than 4095 characters. Such lines were truncated, leading to shortened
3452 data being returned. It should now handle lines of any length.
3454 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3455 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3458 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3459 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3460 no longer happens in comment lines. A list of replacements is now placed
3461 at the head of all of the source files, except those whose only change is
3462 to replace PERL_COMMAND in the very first #! line.
3464 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3465 messages on the queue, with a bottom-up merge sort, using code contributed
3466 by Michael Haardt. This should make operations like -bp somewhat faster on
3467 large queues. It won't affect queue runners, except when queue_run_in_order
3470 51. Installed eximstats 1.31 in the distribution.
3472 52. Added support for SRV lookups to the dnslookup router.
3474 53. If an ACL referred to $message_body or $message_body_end, the value was not
3475 reset for any messages that followed in the same SMTP session.
3477 54. The store-handling optimization for building very long strings was not
3478 differentiating between the different store pools. I don't think this
3479 actually made any difference in practice, but I've tidied it.
3481 55. While running the routers to verify a sender address, $sender_address
3482 was still set to the sender address. This is wrong, because when routing to
3483 send a bounce to the sender, it would be empty. Therefore, I have changed
3484 it so that, while verifying a sender address, $sender_address is set to <>.
3485 (There is no change to what happens when verifying a recipient address.)
3487 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3488 target A or AAAA records (if not already returned) without resetting the
3489 qualify_single or search_parents options of the DNS resolver. These are
3490 inappropriate in this case because the targets of MX and SRV records must
3491 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3492 target that, when qualified, matched something in the local domain. These
3493 two options are now turned off when doing these lookups.
3495 57. It seems that at least some releases of Reiserfs (which does not have the
3496 concept of a fixed number of inodes) returns zero and not -1 for the
3497 number of available inodes. This interacted badly with check_spool_inodes,
3498 which assumed that -1 was the "no such thing" setting. What I have done is
3499 to check that the total number of inodes is greater than zero before doing
3500 the test of how many are available.
3502 58. When a "warn" ACL statement has a log_message modifier, the message is
3503 remembered, and not repeated. This is to avoid a lot of repetition when a
3504 message has many recipients that cause the same warning to be written.
3505 Howewer, Exim was preserving the list of already written lines for an
3506 entire SMTP session, which doesn't seem right. The memory is now reset if a
3507 new message is started.
3509 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3510 debugging output unless log_rewrite was also set.
3512 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3513 of (int)(handle) when we know that handle contains (void *)(-1).
3515 61. The Exim daemon panic-logs an error return when it closes the incoming
3516 connection. However "connection reset by peer" seems to be common, and
3517 isn't really an error worthy of noting specially, so that particular error
3520 62. When Exim is trying to find all the local interfaces, it used to panic and
3521 die if the ioctl to get the interface flags failed. However, it seems that
3522 on at least one OS (Solaris 9) it is possible to have an interface that is
3523 included in the list of interfaces, but for which you get a failure error
3524 for this call. This happens when the interface is not "plumbed" into a
3525 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3526 failure of the "get flags" call assumes that the interface is down.
3528 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3529 makes life easier for people who are doing arithmetic on fields extracted
3530 from dates, where you often get leading zeros that should not be
3531 interpreted as octal.
3533 64. Added qualify_domain to the redirect router, to override the global
3536 65. If a pathologically long header line contained very many addresses (the
3537 report of this problem mentioned 10 000) and each of them was rewritten,
3538 Exim could use up a very large amount of memory. (It kept on making new
3539 copies of the header line as it rewrote, and never released the old ones.)
3540 At the expense of a bit more processing, the header rewriting function has
3541 been changed so that it no longer eats memory in this way.
3543 66. The generation of the Received: header has been moved from the time that a
3544 message starts to be received, to the time that it finishes. The timestamp
3545 in the Received: header should now be very close to that of the <= log
3546 line. There are two side-effects of this change:
3548 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3549 logged header lines no longer include the local Received: line, because
3550 it has not yet been created. The same applies to a copy of the message
3551 that is returned to a non-SMTP sender when a message is rejected.
3553 (b) When a filter file is tested using -bf, no additional Received: header
3554 is added to the test message. After some thought, I decided that this
3557 This change does not affect the value of $received_for. It is still set
3558 after address rewriting, but before local_scan() is called.
3560 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3562 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3563 gave an unhelpful panic error message, and a defer error. I have managed to
3564 change this behaviour so that it now rejects any supplied certificate,
3565 which seems right, as the list of acceptable certificates is empty.
3567 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3568 gave an unhelpful defer error. I have not managed to make this reject any
3569 supplied certificates, but the error message it gives is "no certificate
3570 supplied", which is not helpful.
3572 70. exigrep's output now also includes lines that are not associated with any
3573 message, but which match the given pattern. Implemented by a patch from
3574 Martin Sluka, which also tidied up the Perl a bit.
3576 71. Recipient callout verification, like sender verification, was using <> in
3577 the MAIL FROM command. This isn't really the right thing, since the actual
3578 sender may affect whether the remote host accepts the recipient or not. I
3579 have changed it to use the actual sender in the callout; this means that
3580 the cache record is now keyed on a recipient/sender pair, not just the
3581 recipient address. There doesn't seem to be a real danger of callout loops,
3582 since a callout by the remote host to check the sender would use <>.
3583 [SEE ABOVE: changed after hitting problems.]
3585 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3586 temporary errors. However, in the case of such a code being given after
3587 the end of a data transmission (i.e. after ".") Exim was failing to write
3588 a retry record for the message. (Yes, there was some broken host that was
3589 actually sending 8xx at this point.)
3591 73. An unknown lookup type in a host list could cause Exim to panic-die when
3592 the list was checked. (An example that provoked this was putting <; in the
3593 middle of a list instead of at the start.) If this happened during a DATA
3594 ACL check, a -D file could be left lying around. This kind of configuration
3595 error no longer causes Exim to die; instead it causes a defer errror. The
3596 incident is still logged to the main and panic logs.
3598 74. Buglet left over from Exim 3 conversion. The message "too many messages
3599 in one connection" was written to the rejectlog but not the mainlog, except
3600 when address rewriting (yes!) was being logged.
3602 75. Added write_rejectlog option.
3604 76. When a system filter was run not as root (that is, when system_filter_user
3605 was set), the values of the $n variables were not being returned to the
3606 main process; thus, they were not subsequently available in the $sn
3609 77. Added +return_path_on_delivery log selector.
3611 78. A connection timeout was being treated differently from recipients deferred
3612 when testing hosts_max_try with a message that was older than the host's
3613 retry timeout. (The host should not be counted, thus allowing all hosts to
3614 be tried at least once before bouncing.) This may have been the cause of an
3615 occasionally reported bug whereby a message would remain on the queue
3616 longer than the retry timeout, but would be bounced if a delivery was
3617 forced. I say "may" because I never totally pinned down the problem;
3618 setting up timeout/retry tests is difficult. See also the next item.
3620 79. The ultimate address timeout was not being applied to errors that involved
3621 a combination of host plus message (for example, a timeout on a MAIL
3622 command). When an address resolved to a number of possible hosts, and they
3623 were not all tried for each delivery (e.g. because of hosts_max_try), a
3624 message could remain on the queue longer than the retry timeout.
3626 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3629 81. Fixed an obscure SMTP outgoing bug which required at least the following
3630 conditions: (a) there was another message waiting for the same server;
3631 (b) the server returned 5xx to all RCPT commands in the first message so
3632 that the message was not completed; (c) the server dropped the connection
3633 or gave a negative response to the RSET that Exim sends to abort the
3634 transaction. The observed case was a dropped connection after DATA that had
3635 been sent in pipelining mode. That is, the server had advertised PIPELINING
3636 but was not implementing it correctly. The effect of the bug was incorrect
3637 behaviour, such as trying another host, and this could lead to a crash.
3643 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3644 and daemon.c were passed as pointers to ints; they should have been
3645 pointers to socklen_t variables (which are typically unsigned ints).
3647 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3650 3. Fixed a really odd bug that affected only the testing scheme; patching a
3651 certain fixed string in the binary changed the value of another string that
3652 happened to be identical to the end of the original first string.
3654 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3655 name", it returns that address as the IP address. On some operating
3656 systems (e.g. Solaris), it also passes back the IP address string as the
3657 "host name". However, on others (e.g. Linux), it passes back an empty
3658 string. Exim wasn't checking for this, and was changing the host name to an
3659 empty string, assuming it had been canonicized.
3661 5. Although rare, it is permitted to have more than one PTR record for a given
3662 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3663 all the names associated with an address, because they do in Solaris.
3664 However, it seems that they do not in Linux for data that comes from the
3665 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3666 I found this out when I moved to a new Linux workstation and tried to run
3667 the Exim test suite.
3669 To get round this problem I have changed the code so that it now does its
3670 own call to the DNS to look up PTR records when searching for a host name.
3671 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3672 addresses that are only in /etc/hosts are still found.
3674 This behaviour is, however, controlled by an option called host_lookup_
3675 order, which defaults to "bydns:byaddr". If people want to use the other
3676 order, or indeed, just use one or the other means of lookup, they can
3677 specify it in this variable.
3679 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3680 some operating systems, this comes back from gethostbyaddr() as an empty
3681 string, and this is what Exim used to test for. However, it seems that in
3682 other systems, "." is yielded. Exim now tests for this case too.
3684 7. The values of check_spool_space and check_log_space are now held internally
3685 as a number of kilobytes instead of an absolute number of bytes. If a
3686 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3687 kilobyte. This means that much larger values can be stored.
3689 8. Exim monitor: an attempt to get the action menu when not actually pointing
3690 at a message produces an empty menu entitled "No message selected". This
3691 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3692 no entries in it ("Shell widget menu has zero width and/or height"). So I
3693 have added a single, blank menu entry in this case.
3695 9. Added ${quote_local_part.
3697 10. MIME decoding is now applied to the contents of Subject: header lines when
3700 11. Now that a reference to $sender_host_address automatically causes a reverse
3701 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3702 host lookup before query-style lookups in lists that might use this
3703 variable. This has therefore been abolished, and the "net-" prefix is no
3704 longer necessary for query-style lookups.
3706 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3707 have been a typo for LFLAGS, so it has been changed.
3709 13. The install script calls Exim with "-C /dev/null" in order to find the
3710 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3711 to be output. Howeve, since Exim outputs its version number before the
3712 error, it didn't break the script. It just looked ugly. I fixed this by
3713 always allowing "-C /dev/null" if the caller is root.
3715 14. Ignore overlarge ACL variable number when reading spool file - insurance
3716 against a later release with more variables having written the file.
3718 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3719 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3720 this, as well as the form without the "IPv6" on the front.
3722 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3723 OS/Makefile-Darwin file.
3725 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3726 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3728 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3729 variable. However, some OS use ints for inodes. I've added cast to long int
3730 to get rid of the compiler warning.
3732 19. I had forgotten to lock out "/../" in configuration file names when
3733 ALT_CONFIG_PREFIX was set.
3735 20. Routers used for verification do not need to specify transports. However,
3736 if such a router generated a host list, and callout was configured, Exim
3737 crashed, because it could not find a port number from the (non-existent)
3738 transport. It now assumes port 25 in this circumstance.
3740 21. Added the -t option to exigrep.
3742 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3743 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3744 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3745 them is set, LOOKUP_LSEARCH is forced.
3747 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3749 24. Added sender and host information to the "rejected by local_scan()" log
3750 line; previously there was no indication of these.
3752 25. Added .include_if_exists.
3754 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3755 Linux. It turns out that not all file systems support this. Apparently some
3756 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3757 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3758 not supported on the file descriptor, is now ignored when Exim is trying to
3759 sync a directory. This applies only to Linux.
3761 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3763 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3764 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3767 29. Added support for maildirsize files from supplied patch (modified a bit).
3769 30. The use of :fail: followed by an empty string could lead Exim to respond to
3770 sender verification failures with (e.g.):
3772 550 Verification failed for <xxx>
3773 550 Sender verify failed
3775 where the first response line was missing the '-' that indicates it is not
3776 the final line of the response.
3778 31. The loop for finding the name of the user that called Exim had a hardwired
3779 limit of 10; it now uses the value of finduser_retries, which is used for
3780 all other user lookups.
3782 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3785 33. Exim was neglecting to zero errno before one call of strtol() when
3786 expanding a string and expecting an integer value. On some systems this
3787 resulted in spurious "integer overflow" errors. Also, it was casting the
3788 result into an int without checking.
3790 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3791 did not work. The code looks as if it has *never* worked, though it appears
3792 to have been documented since at least releast 1.62. I have made it work.
3794 35. The "timeout_DNS" error in retry rules, also documented since at least
3795 1.62, also never worked. As it isn't clear exactly what this means, and
3796 clearly it isn't a major issue, I have abolished the feature by treating it
3797 as "timeout", and writing a warning to the main and panic logs.
3799 36. The display of retry rules for -brt wasn't always showing the error code
3802 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3803 timeout_connect_A, timeout_connect_MX.
3805 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3806 to the empty sender.
3808 39. The daemon was not analysing the content of -oX till after it had closed
3809 stderr and disconnected from the controlling terminal. This meant that any
3810 syntax errors were only noted on the panic log, and the return code from
3811 the command was 0. By re-arranging the code a little, I've made the
3812 decoding happen first, so such errors now appear on stderr, and the return
3813 code is 1. However, the actual setting up of the sockets still happens in
3814 the disconnected process, so errors there are still only recorded on the
3817 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3818 connections (as happens on some IP stacks) was logged at start up time as
3819 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3820 it from "IPv6 and IPv4", which means that two separate sockets are being
3823 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3824 says whether AF_INET or AF_INET6 was passed as an argument.
3826 42. Exiwhat output was messed up when time zones were included in log
3829 43. Exiwhat now gives more information about the daemon's listening ports,
3830 and whether -tls-on-connect was used.
3832 44. The "port" option of the smtp transport is now expanded.
3834 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3835 silently ignored. Now an error message is written to the main and panic
3838 46. There's a new ACL modifier called "logwrite" which writes to a log file
3839 as soon as it is encountered.
3841 47. Added $local_user_uid and $local_user_gid at routing time.
3843 48. Exim crashed when trying to verify a sender address that was being
3846 49. Exim was recognizing only a space character after ".include". It now also
3847 recognizes a tab character.
3849 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3850 extracting the relevant information from the specification. The man page no
3851 longer contains scrambled data for the -d option, and I've added a section
3852 at the front about calling Exim under different names.
3854 51. Added "extra_headers" argument to the "mail" command in filter files.
3856 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3859 53. Installed eximstats 1.29.
3861 54. Added transport_filter_timeout as a generic transport option.
3863 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3864 Cc: header lines. This was required by RFC 822, but it not required by RFC
3867 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3868 incoming messages that did not have them. Now it does so only if the
3869 message originates locally, that is, if there is no associated remote host
3870 address. When Resent- header lines are present, this applies to the Resent-
3871 lines rather than the non-Resent- lines.
3873 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3874 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3876 58. Messages for configuration errors now include the name of the main
3877 configuration file - useful now that there may be more than one file in a
3878 list (.included file names were always shown).
3880 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3881 for those rare installations that do not start the daemon as root or run it
3882 setuid root. I've cut out the call to initgroups() if the daemon is not
3885 60. The Exim user and group can now be bound into the binary as text strings
3886 that are looked up at the start of Exim's processing.
3888 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3890 62. Added $mailstore_basename variable.
3892 63. Installed patch to sieve.c from Michael Haardt.
3894 64. When Exim failed to open the panic log after failing to open the main log,
3895 the original message it was trying to log was written to stderr and debug
3896 output, but if they were not available (the usual case in production), it
3897 was lost. Now it is written to syslog before the two lines that record the
3898 failures to open the logs.
3900 65. Users' Exim filters run in subprocesses under the user's uid. It is
3901 possible for a "deliver" command or an alias in a "personal" command to
3902 provoke an address rewrite. If logging of address rewriting is configured,
3903 this fails because the process is not running as root or exim. There may be
3904 a better way of dealing with this, but for the moment (because 4.30 needs
3905 to be released), I have disabled address rewrite logging when running a
3906 filter in a non-root, non-exim process.
3912 1. The buildconfig auxiliary program wasn't quoting the value set for
3913 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3914 not defined. This bug was masked in 4.22 by the effect that was fixed in
3917 2. Some messages that were rejected after a message id was allocated were
3918 shown as "incomplete" by exigrep. It no longer does this for messages that
3919 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3921 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3922 have allow_domain_literals set, the ID did not get logged in the <= line.
3923 Domain literals are now always recognized in Message-ID: header lines.
3925 4. The first argument for a ${extract expansion item is the key name or field
3926 number. Leading and trailing spaces in this item were not being ignored,
3927 causing some misleading effects.
3929 5. When deliver_drop_privilege was set, single queue runner processes started
3930 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3931 same command in the process it spins off) were not dropping privilege.
3933 6. When the daemon running as "exim" started a queue runner, it always
3934 re-executed Exim in the spun-off process. This is a waste of effort when
3935 deliver_drop_privilege is set. The new process now just calls the
3936 queue-runner function directly.
3942 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3945 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3946 ignored. Though the use of -oP was forcing the writing of a pid file, it
3947 was always written to the default place.
3949 3. If the message "no IP address found for host xxxx" is generated during
3950 incoming verification, it is now followed by identification of the incoming
3951 connection (so you can more easily find what provoked it).
3953 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3955 5. Added some features to "harden" Exim a bit more against certain attacks:
3957 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3958 be put in Local/Makefile. This is like the never_users runtime option,
3959 but it cannot be overridden. The default setting is "root".
3961 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3962 prefix string with which any file named in a -C command line option
3965 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3966 is retained for -C and -D only if the caller of Exim is root. Without
3967 it, the exim user may also use -C and -D and retain privilege.
3969 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3970 command line option is disabled.
3972 6. Macro names set by the -D option must start with an upper case letter, just
3973 like macro names defined in the configuration file.
3975 7. Added "dereference=" facility to LDAP.
3977 8. Two instances of the typo "uknown" in the source files are fixed.
3979 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3980 the Configure-Makefile script screwed up while processing it.
3982 10. Incorporated PCRE 4.4.
3984 11. The SMTP synchronization check was not operating right at the start of an
3985 SMTP session. For example, it could not catch a HELO sent before the client
3986 waited for the greeting. There is now a check for outstanding input at the
3987 point when the greeting is written. Because of the duplex, asynchronous
3988 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3989 way, but not yet received, when the check is performed.
3991 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3992 on TCP/IP sockets, because this apparently causes some broken clients to
3995 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3996 unchanged) from the Cygwin maintainer.
3998 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3999 was supported instead of testing for mbx. Effectively a typo.
4001 15. The spa authenticator server code was not checking that the input it
4002 received was valid base64.
4004 16. The debug output line for the "set" modifier in ACLs was not showing the
4005 name of the variable that was being set.
4007 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4009 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4010 lookup on its own. Something else was needed to trigger the lookup. For
4011 example, a match in host_lookup or the need for a host name in a host list.
4012 Now, if $sender_host_name is referenced and the host name has not yet been
4013 looked up, a lookup is performed. If the lookup fails, the variable remains
4014 empty, and $host_lookup_failed is set to "1".
4016 19. Added "eqi" as a case-independent comparison operator.
4018 20. The saslauthd authentication condition could segfault if neither service
4019 nor realm was specified.
4021 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4022 error message that was logged was misleading, and incoming SMTP
4023 connections were dropped. The message is now more accurate, and temporary
4024 errors are given to SMTP connections.
4026 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4027 (except RSET) with a 421 error, until QUIT is received. However, it was
4028 failing to send a response to QUIT.
4030 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4031 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4032 ACL is now run after the helo_try_verify_hosts code.
4034 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4035 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4036 case-independent, so other case variants are also recognized). Apparently
4037 some systems use these upper case variants.
4039 25. If more than two messages were waiting for the same host, and a transport
4040 filter was specified for the transport, Exim sent two messages over the
4041 same TCP/IP connection, and then failed with "socket operation on non-
4042 socket" when it tried to send the third.
4044 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4046 27. The extern definition of crypt16() in expand.c was not being excluded when
4047 the OS had its own crypt16() function.
4049 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4050 as a preferred synonym for return_size_limit, both as an option and as an
4053 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4055 30. Changed the default configuration ACL to relax the local part checking rule
4056 for addresses that are not in any local domains. For these addresses,
4057 slashes and pipe symbols are allowed within local parts, but the sequence
4058 /../ is explicitly forbidden.
4060 31. SPA server authentication was not clearing the challenge buffer before
4063 32. log_message in a "warn" ACL statement was writing to the reject log as
4064 well as to the main log, which contradicts the documentation and doesn't
4065 seem right (because no rejection is happening). So I have stopped it.
4067 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4068 However, I am unable to do any testing of this.
4070 34. Fixed an infelicity in the appendfile transport. When checking directories
4071 for a mailbox, to see if any needed to be created, it was accidentally
4072 using path names with one or more superfluous leading slashes; tracing
4073 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4075 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4076 contents are added to the log line that is written for every discarded
4077 recipient. (Previously a log_message setting was ignored.)
4079 36. The ${quote: operator now quotes the string if it is empty.
4081 37. The install script runs exim in order to find its version number. If for
4082 some reason other than non-existence or emptiness, which it checks, it
4083 could not run './exim', it was installing it with an empty version number,
4084 i.e. as "exim-". This error state is now caught, and the installation is
4087 38. An argument was missing from the function that creates an error message
4088 when Exim fails to connect to the socket for saslauthd authentication.
4089 This could cause Exim to crash, or give a corrupted message.
4091 39. Added isip, isip4, and isip6 to ${if conditions.
4093 40. The ACL variables $acl_xx are now saved with the message, and can be
4094 accessed later in routers, transports, and filters.
4096 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4097 strings in the file are not string-expanded.
4099 42. If a MAIL command specified a SIZE value that was too large to fit into an
4100 int variable, the check against message_size_limit failed. Such values are
4101 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4102 day this will have to be increased, but I don't think I want to be around
4103 when emails are that large.
4110 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4111 iconv() is not standard in FreeBSD.
4113 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4114 IPv6 enabled. The observed symptom was a segmentation fault on return from
4115 the function os_common_find_running_interfaces() in src/os.c.
4117 3. In the check_special_case() function in daemon.c I had used "errno" as an
4118 argument name, which causes warnings on some systems. This was basically a
4119 typo, since it was named "eno" in the comments!
4121 4. The code that waits for the clock to tick (at a resolution of some fraction
4122 of a second) so as to ensure message-id uniqueness was always waiting for
4123 at least one whole tick, when it could have waited for less. [This is
4124 almost certainly not relevant at current processor speeds, where it is
4125 unlikely to ever wait at all. But we try to future-proof.]
4127 5. The function that sleeps for a time interval that includes fractions of a
4128 second contained a race. It did not block SIGALRM between setting the
4129 timer, and suspending (a couple of lines later). If the interval was short
4130 and the sigsuspend() was delayed until after it had expired, the suspension
4131 never ended. On busy systems this could lead to processes getting stuck for
4134 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4135 process, before it forks any delivery processes. The open lookup caching
4136 mechanism meant that the open file or database connection was passed into
4137 the delivery process. The problem was that delivery processes always tidy
4138 up cached lookup data. This could cause a problem for the next delivery
4139 process started by the queue runner, because the external queue runner
4140 process does not know about the closure. So the next delivery process
4141 still has data in the lookup cache. In the case of a file lookup, there was
4142 no problem because closing a file descriptor in a subprocess doesn't affect
4143 the parent. However, if the lookup was caching a connection to a database,
4144 the connection was closed, and the second delivery process was likely to
4145 see errors such as "PGSQL: query failed: server closed the connection
4146 unexpectedly". The problem has been fixed by closing all cached lookups
4147 in a queue runner before running a delivery process.
4149 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4150 seem to have the "const" qualifier which it has on other OS. I've
4153 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4154 *of the same type* (client or server) with the same public name that an
4155 error should be diagnosed.
4157 9. When Exim looked up a host name for an IP address, but failed to find the
4158 original IP address when looking up the host name (a safety check), it
4159 output the message "<ip address> does not match any IP for NULL", which was
4160 confusing, to say the least. The bug was that the host name should have
4161 appeared instead of "NULL".
4163 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4164 user that is built into the binary, and the -C or -D options is used, root
4165 privilege is dropped before the configuration file is read. In addition,
4166 logging is switched to stderr instead of the normal log files. If the
4167 configuration then re-defines the Exim user, the unprivileged environment
4168 is probably not what is expected, so Exim logs a panic warning message (but
4171 However, if deliver_drop_privilege is set, the unprivileged state may well
4172 be exactly what is intended, so the warning has been cut out in that case,
4173 and Exim is allowed to try to write to its normal log files.
4179 1. smtp_return_error_details was not giving details for temporary sender
4180 or receiver verification errors.
4182 2. Diagnose a configuration error if two authenticators have the same public
4185 3. Exim used not to create the message log file for a message until the first
4186 delivery attempt. This could be confusing when incoming messages were held
4187 for policy or load reasons. The message log file is now created at the time
4188 the message is received, and an initial "Received" line is written to it.
4190 4. The automatically generated man page for command line options had a minor
4191 bug that caused no ill effects; however, a more serious problem was that
4192 the procedure for building the man page automatically didn't always
4193 operate. Consequently, release 4.20 contains an out-of-date version. This
4194 shouldn't happen again.
4196 5. When building Exim with embedded Perl support, the script that builds the
4197 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4198 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4200 6. The freeze_tell option was not being used for messages that were frozen on
4201 arrival, either by an ACL or by local_scan().
4203 7. Added the smtp_incomplete_transaction log selector.
4205 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4206 was accepting AUTH without a new EHLO.
4208 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4209 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4212 10. Logging of TCP/IP connections (when configured) now happens in the main
4213 daemon process instead of the child process, so that the TCP/IP connection
4214 count is more accurate (but it can never be perfect).
4216 11. The use of "drop" in a nested ACL was not being handled correctly in the
4217 outer ACL. Now, if condition failure induced by the nested "drop" causes
4218 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4219 or "require"), the connection is dropped.
4221 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4222 that yield "discard" can now be used with an "accept" or a "discard" verb,
4223 but an error is generated for any others (because I can't see a useful way
4224 to define what should happen).
4226 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4227 are now processed in the same way as lines in the Exim configuration file.
4228 In particular, continuation lines are supported.
4230 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4232 15. Added -ti meaning -t -i.
4234 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4235 domains, and warn by logging if others are found.
4237 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4238 returned from GIFCONF when Exim is trying to find the list of interfaces on
4239 a host. The code in os.c has been modified to copy each ifreq to an aligned
4240 structure in all cases.
4242 Also, in some cases, the returned ifreq's were being copied to a 'struct
4243 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4244 means the last couple of bytes of an IPv6 address could be chopped if the
4245 ifreq contained only a normal sockaddr (14 bytes storage).
4247 18. Named domain lists were not supported in the hosts_treat_as_local option.
4248 An entry such as +xxxx was not recognized, and was treated as a literal
4251 19. Ensure that header lines added by a DATA ACL are included in the reject log
4252 if the ACL subsequently rejects the message.
4254 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4255 MD5 (which is deprecated).
4257 21. When testing a filter file using -bf, Exim was writing a message when it
4258 took the sender from a "From " line in the message, but it was not doing so
4259 when it took $return_path from a Return-Path: header line. It now does.
4261 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4262 with a valid header line field name (a series of printing characters
4263 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4265 23. Changed "disc" in the source to "disk" to conform to the documentation and
4266 the book and for uniformity.
4268 24. Ignore Sendmail's -Ooption=value command line item.
4270 25. When execve() failed while trying to run a command in a pipe transport,
4271 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4272 could be confused with a return value of 69 from the command itself. This
4273 has been changed to 127, the value the shell returns if it is asked to run
4274 a non-existent command. The wording for the related log line suggests a
4275 non-existent command as the problem.
4277 26. If received_header_text expands to an empty string, do not add a Received:
4278 header line to the message. (Well, it adds a token one on the spool, but
4279 marks it "old" so that it doesn't get used or transmitted.)
4281 27. Installed eximstats 1.28 (addition of -nt option).
4283 28. There was no check for failure on the call to getsockname() in the daemon
4284 code. This can fail if there is a shortage of resources on the system, with
4285 ENOMEM, for example. A temporary error is now given on failure.
4287 29. Contrary to the C standard, it seems that in some environments, the
4288 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4289 program. Exim now does this explicitly; it affects the formatting of
4290 timestamps using strftime().
4292 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4293 complaints. I've now initialized all the variables, to avoid this.
4295 32. Header lines added by a system filter were not being "seen" during
4296 transport-time rewrites.
4298 33. The info_callback() function passed to OpenSSL is set up with type void
4299 (*)(SSL *, int, int), as described somewhere. However, when calling the
4300 function (actually a macro) that sets it up, the type void(*)() is
4301 expected. I've put in a cast to prevent warnings from picky compilers.
4303 34. If a DNS black list lookup found a CNAME record, but there were no A
4304 records associated with the domain it pointed at, Exim crashed.
4306 35. If a DNS black list lookup returned more than one A record, Exim ignored
4307 all but the first. It now scans all returned addresses if a particular IP
4308 value is being sought. In this situation, the contents of the
4309 $dnslist_value variable are a list of all the addresses, separated by a
4312 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4313 to accept a host name and all its aliases if the forward lookup for any of
4314 them yielded the IP address of the incoming connection. Now it accepts only
4315 those names whose forward lookup yields the correct IP address. Any other
4316 names are discarded. This closes a loophole whereby a rogue DNS
4317 administrator could create reverse DNS records to break through a
4318 wildcarded host restriction in an ACL.
4320 37. If a user filter or a system filter that ran in a subprocess used any of
4321 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4322 the wrong values were passed to the pipe command ($thisaddress had the
4323 value of $0, $0 had the value of $1, etc). This bug was introduced by
4324 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4326 38. Improved the line breaking for long SMTP error messages from ACLs.
4327 Previously, if there was no break point between 40 and 75 characters, Exim
4328 left the rest of the message alone. Two changes have been made: (a) I've
4329 reduced the minimum length to 35 characters; (b) if it can't find a break
4330 point between 35 and 75 characters, it looks ahead and uses the first one
4331 that it finds. This may give the occasional overlong line, but at least the
4332 remaining text gets split now.
4334 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4335 file descriptors might be low, and that setting 1000 would always raise it.
4336 It turns out that in some environments, the limit is already over 1000 and
4337 that lowering it causes trouble. So now Exim takes care not to decrease it.
4339 40. When delivering a message, the value of $return_path is set to $sender_
4340 address at the start of routing (routers may change the value). By an
4341 oversight, this default was not being set up when an address was tested by
4342 -bt or -bv, which affected the outcome if any router or filter referred to
4345 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4346 log only when "message" or "log_message" are set. However, if one of the
4347 conditions was an address verification, or a call to a nested ACL, the
4348 messages generated by the underlying test were being passed through. This
4349 no longer happens. The underlying message is available in $acl_verify_
4350 message for both "message" and "log_message" expansions, so it can be
4351 passed through if needed.
4353 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4354 new expansion $bh_ to give the encoded byte string without charset
4355 translation. Translation happens only if iconv() is available; HAVE_ICONV
4356 indicates this at build time. HEADERS_CHARSET gives the charset to
4357 translate to; headers_charset can change it in the configuration, and
4358 "headers charset" can change it in an individual filter file.
4360 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4361 that creates RFC 2047 encoded "words" labels them as that charset instead
4362 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4363 expansion operator; (ii) when Exim creates a From: line for a local
4364 message; (iii) when a header line is rewritten to include a "phrase" part.
4366 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4367 buggy, causing it to skip the first lines of messages whose message ID
4368 ended in 'D'. This would not have bitten before Exim release 4.14, because
4369 message IDs were unlikely to end in 'D' before then. The effect was to have
4370 incorrect size information for certain domains.
4372 45. #include "config.h" was missing at the start of the crypt16.c module. This
4373 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4376 46. If there was a timeout during a "random" callout check, Exim treated it as
4377 a failure of the random address, and carried on sending RSET and the real
4378 address. If the delay was just some slowness somewhere, the response to the
4379 original RCPT would be taken as a response to RSET and so on, causing
4380 mayhem of various kinds.
4382 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4383 when I implemented it. It didn't allow for the fact that some option values
4384 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4385 the right test for positive values.
4387 48. Domain names in DNS records are case-independent. Exim always looks them up
4388 in lower case. Some resolvers return domain names in exactly the case they
4389 appear in the zone file, that is, they may contain uppercase letters. Not
4390 all resolvers do this - some return always lower case. Exim was treating a
4391 change of case by a resolver as a change of domain, similar to a widening
4392 of a domain abbreviation. This triggered its re-routing code and so it was
4393 trying to route what was effectively the same domain again. This normally
4394 caused routing to fail (because the router wouldn't handle the domain
4395 twice). Now Exim checks for this case specially, and just changes the
4396 casing of the domain that it ultimately uses when it transmits the message
4399 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4402 50. If a filter generated a file delivery with a non-absolute name (possible if
4403 no home directory exists for the router), the forbid_file option was not
4406 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4407 the existing equality matching.
4409 52. Exim was using ints instead of ino_t variables in some places where it was
4410 dealing with inode numbers.
4412 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4413 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4414 called TMPDIR, and if it finds it is different, it changes its value.
4416 54. The smtp_printf() function is now made available to local_scan() so
4417 additional output lines can be written before returning. There is also an
4418 smtp_fflush() function to enable the detection of a dropped connection.
4419 The variables smtp_input and smtp_batched_input are exported to
4422 55. Changed the default runtime configuration: the message "Unknown user"
4423 has been removed from the ACL, and instead placed on the localuser router,
4424 using the cannot_route_message feature. This means that any verification
4425 failures that generate their own messages won't get overridden. Similarly,
4426 the "Unrouteable address" message that was in the ACL for unverifiable
4427 relay addresses has also been removed.
4429 56. Added hosts_avoid_esmtp to the smtp transport.
4431 57. The exicyclog script was not checking for the esoteric option
4432 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4433 will work only if exicyclog is run under the appropriate euid.
4435 58. Following a discussion on the list, the rules by which Exim recognises line
4436 endings on incoming messages have been changed. The -dropcr and drop_cr
4437 options are now no-ops, retained only for backwards compatibility. The
4438 following line terminators are recognized: LF CRLF CR. However, special
4439 processing applies to CR:
4441 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4442 nor a local message in the state where . is a terminator.
4444 (ii) If a bare CR is encountered in a header line, an extra space is added
4445 after the line terminator so as not to end the header. The reasoning
4446 behind this is that bare CRs in header lines are most likely either
4447 to be mistakes, or people trying to play silly games.
4449 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4450 was being incorrectly given as 18 bytes larger than it should have been.
4451 This is a VOB (very old bug).
4453 60. This may never have affected anything current, but just in case it has:
4454 When the local host is found other than at the start of a list of hosts,
4455 the local host, those with the same MX, and any that follow, are discarded.
4456 When the list in question was part of a longer list of hosts, the following
4457 hosts (not currently being processed) were also being discarded. This no
4458 longer happens. I'm not sure if this situation could ever has previously
4461 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4464 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4465 header. This is something that is recommended in a new Internet Draft, and
4466 is something that is documented as being done by Sendmail. There are two
4467 possible values. For messages generated by the autoreply transport, Exim
4470 Auto-Submitted: auto-replied
4472 whereas for all other generated messages (e.g. bounces) it adds
4474 Auto-Submitted: auto-generated
4476 63. The "personal" condition in filters now includes a test for the
4477 Auto-Submitted: header. If it contains the string "auto-" the message it
4478 not considered personal.
4480 64. Added rcpt_include_affixes as a generic transport option.
4482 65. Added queue_only_override (default true).
4484 66. Added the syslog_duplication option.
4486 67. If what should have been the first header line of a message consisted of
4487 a space followed by a colon, Exim was mis-interpreting it as a header line.
4488 It isn't of course - it is syntactically invalid and should therefore be
4489 treated as the start of the message body. The misbehaviour could have
4490 caused a number of strange effects, including loss of data in subsequent
4491 header lines, and spool format errors.
4493 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4494 client host had authenticated. This control can now be exercised by an ACL
4495 for more flexibility.
4497 69. By default, callouts do not happen when testing with -bh. There is now a
4498 variant, -bhc, which does actually run the callout code, including
4499 consulting and updating the callout cache.
4501 70. Added support for saslauthd authentication, courtesy of Alexander
4504 71. If statvfs() failed on the spool or log directories while checking their
4505 size for availability, Exim confusingly gave the error "space shortage".
4506 Furthermore, in debugging mode it crashed with a floating point exception.
4507 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4508 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4509 serious problem, Exim now writes to the main and panic logs when this
4510 happens, with details of the failure. It then refuses to accept the
4511 incoming message, giving the message "spool directory problem" or "log
4512 directory problem" with a 421 code for SMTP messages.
4514 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4515 0, 1, and 2 exist, because some OS complain for execs without them (see
4516 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4517 descriptors. However, the code omitted to check that the open succeeded,
4518 causing mysterious errors if for some reason the permissions on /dev/null
4519 got screwed. Now Exim writes a message to the main and panic logs, and
4520 bombs out if it can't open /dev/null.
4522 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4523 interact so that it is all more flexible. It is supposed to remain
4524 backwards compatible. Also added extra_local_interfaces.
4526 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4527 to bomb out with an assertion failure - to the client this appears as a
4528 connection drop. This problem occurs in the part of the code that was taken
4529 from the Samba project. Fortunately, the assertion is in a very simple
4530 function, so I have fixed this by reproducing the function inline in the
4531 one place where it is called, and arranging for authentication to fail
4532 instead of killing the process with assert().
4534 75. The SPA client code was not working when the server requested OEM rather
4535 than Unicode encoding.
4537 76. Added code to make require_files with a specific uid setting more usable in
4538 the case where statting the file as root fails - usually a non-root-mounted
4539 NFS file system. When this happens and the failure is EACCES, Exim now
4540 forks a subprocess and does the per-uid checking as the relevant uid.
4542 77. Added process_log_path.
4544 78. If log_file_path was not explicitly set, a setting of check_log_space or
4545 check_log_inodes was ignored.
4547 79. If a space check for the spool or log partitions fails, the incident is now
4548 logged. Of course, in the latter case the data may get lost...
4550 80. Added the %p formatting code to string_format() so that it can be used to
4551 print addresses in debug_print(). Adjusted all the address printing in the
4552 debugging in store.c to use %p rather than %d.
4554 81. There was a concern that a line of code in smtp_in.c could overflow a
4555 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4556 initially expressed, the concern was not well-founded, because trailing
4557 spaces are removed early. However, if the trailing spaces were followed by
4558 a NULL, they did not get removed, so the overflow was possible. Two fixes
4561 (a) I re-wrote the offending code in a cleaner fashion.
4562 (b) If an incoming SMTP command contains a NULL character, it is rejected
4565 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4566 runs initgroups(), so that if the Exim user is in any additional groups,
4567 they will be used during message reception.
4573 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob