1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.458 2007/01/23 11:01:09 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.
48 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
49 fixed by 4.65/MH/01 (is this a record?) are fixed:
51 (i) An empty string was always treated as zero by the numeric comparison
52 operators. This behaviour has been restored.
54 (ii) It is documented that the numeric comparison operators always treat
55 their arguments as decimal numbers. This was broken in that numbers
56 starting with 0 were being interpreted as octal.
58 While fixing these problems I realized that there was another issue that
59 hadn't been noticed. Values of message_size_limit (both the global option
60 and the transport option) were treated as octal if they started with 0.
61 The documentation was vague. These values are now always treated as
62 decimal, and I will make that clear in the documentation.
68 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
69 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
72 MH/01 Don't check that the operands of numeric comparison operators are
73 integers when their expansion is in "skipping" mode (fixes bug
74 introduced by 4.64-PH/07).
76 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
77 child addresses, Exim now panics and dies. Previously, because the count
78 is held in a short int, deliveries were likely to be lost. As such a
79 large number of recipients for a single message is ridiculous
80 (performance will be very, very poor), I have chosen to impose a limit
81 rather than extend the field.
87 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
88 leftover -K file (the existence of which was triggered by #402).
89 While we were at it, introduced process PID as part of the -K
90 filename. This should rule out race conditions when creating
93 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
94 processing considerably. Previous code took too long for large mails,
95 triggering a timeout which in turn triggers #401.
97 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
98 in the DK code in transports.c. sendfile() is not really portable,
99 hence the _LINUX specificness.
101 TF/01 In the add_headers option to the mail command in an Exim filter,
102 there was a bug that Exim would claim a syntax error in any
103 header after the first one which had an odd number of characters
106 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
107 callout verification, Exim cached a "reject" for the entire domain. This
108 is correct for most verifications, but it is not correct for a recipient
109 verification with use_sender or use_postmaster set, because in that case
110 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
111 case of MAIL FROM:<> rejection from other early rejections (e.g.
112 rejection of HELO). When verifying a recipient using a non-null MAIL
113 address, the cache is ignored if it shows MAIL FROM:<> rejection.
114 Whatever the result of the callout, the value of the domain cache is
115 left unchanged (for any other kind of callout, getting as far as trying
116 RCPT means that the domain itself is ok).
118 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
121 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
122 manifest itself as EPIPE rather than ECONNECT. When tidying away a
123 session, the daemon ignores ECONNECT errors and logs others; it now
124 ignores EPIPE as well.
126 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
127 (quoted-printable decoding).
129 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
130 later the small subsequent patch to fix an introduced bug.
132 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
134 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
136 PH/08 An error is now given if message_size_limit is specified negative.
138 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
139 to be given (somewhat) arbitrary names.
141 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
144 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
145 miscellaneous code fixes
147 PH/10 Added the log_reject_target ACL modifier to specify where to log
150 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
151 hostname. This is wrong, because it relates to the incoming message (and
152 probably the interface on which it is arriving) and not to the outgoing
153 callout (which could be using a different interface). This has been
154 changed to use the value of the helo_data option from the smtp transport
155 instead - this is what is used when a message is actually being sent. If
156 there is no remote transport (possible with a router that sets up host
157 addresses), $smtp_active_hostname is used.
159 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
160 tweaks were necessary in order to get it to work (see also 21 below):
161 (a) The code assumed that strncpy() returns a negative number on buffer
162 overflow, which isn't the case. Replaced with Exim's string_format()
164 (b) There were several signed/unsigned issues. I just did the minimum
165 hacking in of casts. There is scope for a larger refactoring.
166 (c) The code used strcasecmp() which is not a standard C function.
167 Replaced with Exim's strcmpic() function.
168 (d) The code set only $1; it now sets $auth1 as well.
169 (e) A simple test gave the error "authentication client didn't specify
170 service in request". It would seem that Dovecot has changed its
171 interface. Fortunately there's a specification; I followed it and
172 changed what the client sends and it appears to be working now.
174 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
177 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
178 address A is aliased to B and C, where B exists and C does not. Without
179 -v the output is "A verified" because verification stops after a
180 successful redirection if more than one address is generated. However,
181 with -v the child addresses are also verified. Exim was outputting "A
182 failed to verify" and then showing the successful verification for C,
183 with its parentage. It now outputs "B failed to verify", showing B's
184 parentage before showing the successful verification of C.
186 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
187 look up a TXT record in a specific list after matching in a combined
190 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
191 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
192 they consult the DNS. I had assumed they would set it the way they
193 wanted; and indeed my experiments on Linux seem to show that in some
194 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
195 To be on the safe side, however, I have now made the interface to
196 host_find_byname() similar to host_find_bydns(), with an argument
197 containing the DNS resolver options. The host_find_byname() function now
198 sets these options at its start, just as host_find_bydns() does. The smtp
199 transport options dns_qualify_single and dns_search_parents are passed to
200 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
201 of host_find_byname() use the default settings of RES_DEFNAMES
202 (qualify_single) but not RES_DNSRCH (search_parents).
204 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
205 spool_read_header() do less string testing, by means of a preliminary
206 switch on the second character of optional "-foo" lines. (This is
207 overdue, caused by the large number of possibilities that now exist.
208 Originally there were few.) While I was there, I also converted the
209 str(n)cmp tests so they don't re-test the leading "-" and the first
210 character, in the hope this might squeeze out yet more improvement.
212 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
213 flag allowing group syntax was set by the header_syntax check but not
214 turned off, possible causing trouble later; (2) The flag was not being
215 set at all for the header_verify test, causing "group"-style headers to
216 be rejected. I have now set it in this case, and also caused header_
217 verify to ignore an empty address taken from a group. While doing this, I
218 came across some other cases where the code for allowing group syntax
219 while scanning a header line wasn't quite right (mostly, not resetting
220 the flag correctly in the right place). These bugs could have caused
221 trouble for malformed header lines. I hope it is now all correct.
223 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
224 with the "reply" argument non-NULL. The code, however (which originally
225 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
226 but it didn't always do it. This confused somebody who was copying the
227 code for some other use. I have removed all the tests.
229 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
230 feature that was used to support insecure browsers during the U.S. crypto
231 embargo. It requires special client support, and Exim is probably the
232 only MTA that supported it -- and would never use it because real RSA is
233 always available. This code has been removed, because it had the bad
234 effect of slowing Exim down by computing (never used) parameters for the
235 RSA_EXPORT functionality.
237 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
238 authenticator to fail if there's a tab character in the incoming data
239 (there should never be unless someone is messing about, as it's supposed
240 to be base64-encoded). Also added, on Timo's advice, the "secured" option
241 if the connection is using TLS or if the remote IP is the same as the
242 local IP, and the "valid-client-cert option" if a client certificate has
245 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
246 authenticators. This can be used for authorization after authentication
247 succeeds. (In the case of plaintext, it servers for both authentication
250 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
251 if any retry times were supplied.
253 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
254 connection, where there is no client IP address to check. In this
255 situation, the verify now always succeeds.
257 PH/25 Applied John Jetmore's -Mset patch.
259 PH/26 Added -bem to be like -Mset, but loading a message from a file.
261 PH/27 In a string expansion for a processed (not raw) header when multiple
262 headers of the same name were present, leading whitespace was being
263 removed from all of them, but trailing whitespace was being removed only
264 from the last one. Now trailing whitespace is removed from each header
265 before concatenation. Completely empty headers in a concatenation (as
268 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
269 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
271 PH/29 [Removed. This was a change that I later backed out, and forgot to
272 correct the ChangeLog entry (that I had efficiently created) before
273 committing the later change.]
275 PH/30 Exim was sometimes attempting to deliver messages that had suffered
276 address errors (4xx response to RCPT) over the same connection as other
277 messages routed to the same hosts. Such deliveries are always "forced",
278 so retry times are not inspected. This resulted in far too many retries
279 for the affected addresses. The effect occurred only when there were more
280 hosts than the hosts_max_try setting in the smtp transport when it had
281 the 4xx errors. Those hosts that it had tried were not added to the list
282 of hosts for which the message was waiting, so if all were tried, there
283 was no problem. Two fixes have been applied:
285 (i) If there are any address or message errors in an SMTP delivery, none
286 of the hosts (tried or untried) are now added to the list of hosts
287 for which the message is waiting, so the message should not be a
288 candidate for sending over the same connection that was used for a
289 successful delivery of some other message. This seems entirely
290 reasonable: after all the message is NOT "waiting for some host".
291 This is so "obvious" that I'm not sure why it wasn't done
292 previously. Hope I haven't missed anything, but it can't do any
293 harm, as the worst effect is to miss an optimization.
295 (ii) If, despite (i), such a delivery is accidentally attempted, the
296 routing retry time is respected, so at least it doesn't keep
297 hammering the server.
299 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
300 in ${readsocket because some servers need this prod.
302 PH/32 Added some extra debug output when updating a wait-xxx database.
304 PH/33 The hint "could be header name not terminated by colon", which has been
305 given for certain expansion errors for a long time, was not being given
306 for the ${if def:h_colon_omitted{... case.
308 PH/34 The spec says: "With one important exception, whenever a domain list is
309 being scanned, $domain contains the subject domain." There was at least
310 one case where this was not true.
312 PH/35 The error "getsockname() failed: connection reset by peer" was being
313 written to the panic log as well as the main log, but it isn't really
314 panic-worthy as it just means the connection died rather early on. I have
315 removed the panic log writing for the ECONNRESET error when getsockname()
318 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
319 runs only) independently of the message's sender address. This meant
320 that, if the 4xx error was in fact related to the sender, a different
321 message to the same recipient with a different sender could confuse
322 things. In particualar, this can happen when sending to a greylisting
323 server, but other circumstances could also provoke similar problems.
324 I have changed the default so that the retry time for these errors is now
325 based a combination of the sender and recipient addresses. This change
326 can be overridden by setting address_retry_include_sender=false in the
329 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
330 remote server are returned as part of bounce messages. This was not
331 happening for LMTP over a pipe (the lmtp transport), but now it is the
332 same for both kinds of LMTP.
334 PH/38 Despite being documented as not happening, Exim was rewriting addresses
335 in header lines that were in fact CNAMEs. This is no longer the case.
337 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
338 and queue runs started by the daemon processed all messages. This has
339 been fixed so that -R and -S can now usefully be given with -q<time>.
341 PH/40 Import PCRE release 6.7 (fixes some bugs).
343 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
345 PH/42 Give an error if -q is specified more than once.
347 PH/43 Renamed the variables $interface_address and $interface_port as
348 $received_ip_address and $received_port, to make it clear that these
349 values apply to message reception, and not to the outgoing interface when
350 a message is delivered. (The old names remain recognized, of course.)
352 PH/44 There was no timeout on the connect() call when using a Unix domain
353 socket in the ${readsocket expansion. There now is.
355 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
356 be meaningful with "accept".
358 SC/01 Eximstats V1.43
359 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
361 SC/02 Eximstats V1.44
362 Use a glob alias rather than an array ref in the generated
363 parser. This improves both readability and performance.
365 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
366 Collect SpamAssassin and rejection statistics.
367 Don't display local sender or destination tables unless
368 there is data to show.
369 Added average volumes into the top table text output.
371 SC/04 Eximstats V1.46
372 Collect data on the number of addresses (recipients)
373 as well as the number of messages.
375 SC/05 Eximstats V1.47
376 Added 'Message too big' to the list of mail rejection
377 reasons (thanks to Marco Gaiarin).
379 SC/06 Eximstats V1.48
380 Mainlog lines which have GMT offsets and are too short to
381 have a flag are now skipped.
383 SC/07 Eximstats V1.49 (Alain Williams)
384 Added the -emptyok flag.
386 SC/08 Eximstats V1.50
387 Fixes for obtaining the IP address from reject messages.
389 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
390 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
391 whitesspace changes from 4.64-PH/27
393 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
396 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
397 are found, allow negative numbers in numeric criteria)
399 JJ/06 exipick.20061117.2, added new $message_body_missing variable
401 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
402 to match changes made in 4.64-PH/43
404 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
406 PH/47 Put in an explicit test for a DNS lookup of an address record where the
407 "domain" is actually an IP address, and force a failure. This locks out
408 those revolvers/nameservers that support "A-for-A" lookups, in
409 contravention of the specifications.
411 PH/48 When a host name was looked up from an IP address, and the subsequent
412 forward lookup of the name timed out, the host name was left in
413 $sender_host_name, contrary to the specification.
415 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
416 restricted to single-key lookups, Exim was not diagnosing an error if
417 * or *@ was used with a query-style lookup.
419 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
421 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
422 long ago, but noone interested enough thought of it. Let's just say that
423 the "1.1" means that there are some new functions that weren't there at
424 some point in the past.
426 PH/51 Error processing for expansion failure of helo_data from an smtp
427 transport during callout processing was broken.
429 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
430 tested/used via the -bh/-bhc/-bs options.
432 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
433 bug, fixed in subsequent PCRE releases).
435 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
436 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
438 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
444 SC/01 Use a glob alias rather than an array ref in eximstats generated
445 parser. This improves both readability and performance.
447 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
448 Don't display local sender or destination tables in eximstats unless
449 there is data to show.
450 Added average volumes into the eximstats top table text output.
452 SC/03 Collect data on the number of addresses (recipients) as well
453 as the number of messages in eximstats.
455 TF/01 Correct an error in the documentation for the redirect router. Exim
456 does (usually) call initgroups() when daemonizing.
458 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
459 with consistent privilege compared to when running as a daemon.
461 TF/03 Note in the spec that $authenticated_id is not set for local
462 submissions from trusted users.
464 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
465 Thanks to Dean Brooks <dean@iglou.com> for the patch.
467 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
468 by adding some example configuration directives to the default
469 configuration file. A little bit of work is required to uncomment the
470 directives and define how usernames and passwords are checked, but
471 there is now a framework to start from.
473 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
474 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
475 without this. I don't know how relevant this is to other LDAP libraries.
477 PH/02 Add the verb name to the "unknown ACL verb" error.
479 PH/03 Magnus Holmgren's patch for filter_prepend_home.
481 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
483 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
484 directory not expanded when it should be if an expanded home directory
485 was set for the address (which is overridden by the transport).
487 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
490 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
491 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
492 because it is too late at that time, and has no effect.
494 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
495 security issue with \' (bugzilla #107). I could not use the
496 PQescapeStringConn() function, because it needs a PGconn value as one of
499 PH/08 When testing addresses using -bt, indicate those final addresses that
500 are duplicates that would not cause an additional delivery. At least one
501 person was confused, thinking that -bt output corresponded to deliveries.
502 (Suppressing duplicates isn't a good idea as you lose the information
503 about possibly different redirections that led to the duplicates.)
505 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
506 systems where poll() doesn't work, in particular OS X.
508 PH/10 Added more information to debugging output for retry time not reached.
510 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
511 operations in malware.c.
513 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
516 PH/13 If write_rejectlog was set false when logging was sent to syslog with
517 syslog_duplication set false, log lines that would normally be written
518 both the the main log and to the reject log were not written to syslog at
521 PH/14 In the default configuration, change the use of "message" in ACL warn
522 statements to "add_header".
524 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
525 not followed by a command (e.g. "seen endif").
527 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
528 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
531 PH/17 Added extra conditions to the default value of delay_warning_condition
535 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
536 { match{$h_precedence:}{(?i)bulk|list|junk} } \
537 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
540 The Auto-Submitted: and various List- headers are standardised, whereas I
541 don't think Precedence: ever was.
543 PH/18 Refactored debugging code in route_finduser() to show more information,
544 in particular, the error code if getpwnam() issues one.
546 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
547 This is apparently needed in addition to the PH/07 change above to avoid
548 any possible encoding problems.
550 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
551 but not after initializing Perl.
553 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
554 output them only if debugging. By default they are written stderr,
555 apparently, which is not desirable.
557 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
560 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
563 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
565 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
566 authenticated or an ident call has been made. Suppress the default
567 values for $authenticated_id and $authenticated_sender (but permit -oMai
568 and -oMas) when testing with -bh.
570 PH/24 Re-jigged the order of the tests in the default configuration so that the
571 tests for valid domains and recipients precede the DNS black list and CSA
572 tests, on the grounds that those ones are more expensive.
574 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
575 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
576 This bug exists in every version of Exim that I still have, right back to
579 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
580 However, an attempt to turn on case-sensitivity in a regex key by
581 including (?-i) didn't work because the subject string was already
582 lowercased, and the effects were non-intuitive. It turns out that a
583 one-line patch can be used to allow (?-i) to work as expected.
589 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
590 other effects) broke the use of negated acl sub-conditions.
592 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
595 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
596 "Deny" causes Exim to reject the incoming connection with a 554 error.
597 Unfortunately, if there is a major crisis, such as a disk failure,
598 tcp-wrappers gives "deny", whereas what one would like would be some
599 kind of temporary error. A kludge has been added to help with this.
600 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
601 554 error is used if errno is still zero or contains ENOENT (which occurs
602 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
605 PH/03 Add -lutil to the default FreeBSD LIBS setting.
607 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
608 errors. Otherwise a message that provokes a temporary error (when other
609 messages do not) can cause a whole host to time out.
611 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
612 addresses were routed directly to files or pipes from a redirect router.
613 File deliveries just didn't batch; pipe deliveries might have suffered
616 PH/06 A failure to get a lock for a hints database would erroneously always say
617 "Failed to get write lock", even when it was really a read lock.
619 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
620 of 0600. This has been changed to use the value of the lockfile_mode
621 option (which defaults to 0600).
623 PH/08 Applied small patch from the Sieve maintainer.
625 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
626 folder from quota calculations, a direct delivery into this folder messed
627 up the contents of the maildirsize file. This was because the regex was
628 used only to exclude .Trash (or whatever) when the size of the mailbox
629 was calculated. There was no check that a delivery was happening into an
630 excluded directory. This bug has been fixed by ignoring all quota
631 processing for deliveries into excluded directories.
633 PH/10 Added the maildirfolder_create_regex option to appendfile.
639 PH/01 The code for finding all the local interface addresses on a FreeBSD
640 system running IPv6 was broken. This may well have applied to all BSD
641 systems, as well as to others that have similar system calls. The broken
642 code found IPv4 interfaces correctly, but gave incorrect values for the
643 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
644 that it would not match correctly against @[] and not recognize the IPv6
647 PH/02 The ipliteral router was not recognizing addresses of the form user@
648 [ipv6:....] because it didn't know about the "ipv6:" prefix.
650 PH/03 Added disable_ipv6.
652 PH/04 Changed $reply_address to use the raw form of the headers instead of the
653 decoded form, because it is most often used to construct To: headers
654 lines in autoreplies, and the decoded form may well be syntactically
655 invalid. However, $reply_address has leading white space removed, and all
656 newlines turned into spaces so that the autoreply transport does not
659 PH/05 If group was specified without a user on a router, and no group or user
660 was specified on a transport, the group from the router was ignored.
662 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
663 for visible compile-time settings that can be used to change these
664 numbers, for those that want even more. Backwards compatibility with old
665 spool files has been maintained. However, going back to a previous Exim
666 release will lost any variables that are in spool files.
668 PH/07 Two small changes when running in the test harness: increase delay when
669 passing a TCP/IP connection to a new process, in case the original
670 process has to generate a bounce, and remove special handling of
671 127.0.0.2 (sic), which is no longer necessary.
673 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
674 be the same on different OS.
676 PH/09 Moved a debug statement in filter processing to avoid a race problem when
679 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
680 whether --show-vars was specified or not
682 JJ/02 exipick: Added support for new ACL variable spool format introduced
685 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
686 syntactically invalid From: or Reply-to: line, and a filter used this to
687 generate an autoreply, and therefore failed to obtain an address for the
688 autoreply, Exim could try to deliver to a non-existent relative file
689 name, causing unrelated and misleading errors. What now happens is that
690 it logs this as a hard delivery error, but does not attempt to create a
693 PH/11 The exinext utility has a -C option for testing purposes, but although
694 the given file was scanned by exinext itself; it wasn't being passed on
695 when Exim was called.
697 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
698 an end-of-file indication when reading a command response.
700 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
701 compiled. In many other places in Exim, IPv6 addresses are always
702 recognized, so I have changed this. It also means that IPv4 domain
703 literals of the form [IPV4:n.n.n.n] are now always recognized.
705 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
706 used if the router is not running as root, for example, when verifying at
707 ACL time, or when using -bh. The debugging output from this situation was
708 non-existent - all you got was a failure to exec. I have made two
711 (a) Failures to set uid/gid, the current directory, or a process leader
712 in a subprocess such as that created by queryprogram now generate
713 suitable debugging ouput when -d is set.
715 (b) The queryprogram router detects when it is not running as root,
716 outputs suitable debugging information if -d is set, and then runs
717 the subprocess without attempting to change uid/gid.
719 PH/15 Minor change to Makefile for building test_host (undocumented testing
722 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
723 additional section of a DNS packet that returns MX or SRV records.
724 Instead, it always explicitly searches for A/AAAA records. This avoids
725 major problems that occur when a DNS server includes only records of one
726 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
727 fixed another bug: if SRV records were looked up and the corresponding
728 address records were *not* found in the additional section, the port
729 values from the SRV records were lost.
731 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
732 using the correct key (the original address) when searching the retry
733 rules in order to find which one to use for generating the retry hint.
735 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
736 adding the default one. Similarly, if it contains a Reply-To: header, the
737 errors_reply_to option, if set, is not used.
739 PH/19 When calculating a retry time, Exim used to measure the "time since
740 failure" by looking at the "first failed" field in the retry record. Now
741 it does not use this if it is later than than the arrival time of the
742 message. Instead it uses the arrival time. This makes for better
743 behaviour in cases where some deliveries succeed, thus re-setting the
744 "first failed" field. An example is a quota failure for a huge message
745 when small messages continue to be delivered. Without this change, the
746 "time since failure" will always be short, possible causing more frequent
747 delivery attempts for the huge message than are intended.
748 [Note: This change was subsequently modified - see PH/04 for 4.62.]
750 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
751 $1, $2, $3) because the numerical variables can be reset during some
752 expansion items (e.g. "match"), thereby losing the authentication data.
754 PH/21 Make -bV show the size of off_t variables so that the test suite can
755 decide whether to run tests for quotas > 2G.
757 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
758 mailbox_size, and mailbox_filecount in the appendfile transport. If a
759 filecount value is greater than 2G or if a quota value is greater than 2G
760 on a system where the size of off_t is not greater than 4, a panic error
763 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
764 never match. The debug and -bh output now contains an explicit error
765 message indicating a malformed IPv4 address or mask.
767 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
768 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
771 PH/25 Do not write to syslog when running in the test harness. The only
772 occasion when this arises is a failure to open the main or panic logs
773 (for which there is an explicit test).
775 PH/26 Added the /no_tell option to "control=freeze".
777 PH/27 If a host name lookup failed very early in a connection, for example, if
778 the IP address matched host_lookup and the reverse lookup yielded a name
779 that did not have a forward lookup, an error message of the form "no IP
780 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
781 could be logged. Now it outputs the IP address instead of "NULL".
783 PH/28 An enabling patch from MH: add new function child_open_exim2() which
784 allows the sender and the authenticated sender to be set when
785 submitting a message from within Exim. Since child_open_exim() is
786 documented for local_scan(), the new function should be too.
788 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
789 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
790 results in an empty string is now treated as unset.
792 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
794 PH/31 Added sender_verify_fail logging option.
796 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
797 needed by RFC 822 but not by RFC 2822 was commented out. I have now
798 tidied the source and removed it altogether.
800 PH/33 When a queue run was abandoned because the load average was too high, a
801 log line was always written; now it is written only if the queue_run log
802 selector is set. In addition, the log line for abandonment now contains
803 information about the queue run such as the pid. This is always present
804 in "start" and "stop" lines but was omitted from the "abandon" line.
806 PH/34 Omit spaces between a header name and the colon in the error message that
807 is given when verify = headers_syntax fails (if there are lots of them,
808 the message gets confusing).
810 PH/35 Change the default for dns_check_names_pattern to allow slashes within
811 names, as there are now some PTR records that contain slashes. This check
812 is only to protect against broken name servers that fall over on strange
813 characters, so the fact that it applies to all lookups doesn't matter.
815 PH/36 Now that the new test suite is complete, we can remove some of the
816 special code in Exim that was needed for the old test suite. For example,
817 sorting DNS records because real resolvers return them in an arbitrary
818 order. The new test suite's fake resolver always returns records in the
821 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
822 bounces) except when queue_only is set, to avoid logging races between
823 the different processes.
825 PH/38 Panic-die if .include specifies a non-absolute path.
827 PH/39 A tweak to the "H" retry rule from its user.
829 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
830 a label. They prevented compilation on older perls.
832 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
833 a warning to be raised on newish perls.
835 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
836 on queue. Changes to match documented behaviour of showing count of
837 messages matching specified criteria.
839 PH/40 Changed the default ident timeout from 30s to 5s.
841 PH/41 Added support for the use of login_cap features, on those BSD systems
842 that have them, for controlling the resources used by pipe deliveries.
844 PH/42 The content-scanning code uses fopen() to create files in which to put
845 message data. Previously it was not paying any attention to the mode of
846 the files. Exim runs with umask(0) because the rest of the code creates
847 files with open(), and sets the required mode explicitly. Thus, these
848 files were ending up world-writeable. This was not a big issue, because,
849 being within the spool directory, they were not world-accessible. I have
850 created a function called modefopen, which takes an additional mode
851 argument. It sets umask(777), creates the file, chmods it to the required
852 mode, then resets the umask. All the relevant calls to fopen() in the
853 content scanning code have been changed to use this function.
855 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
856 to 24 hours. This avoids potential overflow problems when processing G
857 and H retry rules. I suspect nobody ever tinkers with this value.
859 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
861 PH/45 When the plaintext authenticator is running as a client, the server's
862 challenges are checked to ensure they are valid base64 strings. By
863 default, the authentication attempt is cancelled if an invalid string is
864 received. Setting client_ignore_invalid_base64 true ignores these errors.
865 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
866 they are received. Thus, the responses can be made to depend on the
867 challenges. If an invalid string is ignored, an empty string is placed in
870 PH/46 Messages that are created by the autoreply transport now contains a
871 References: header, in accordance with RFCs 2822 and 3834.
873 PH/47 Added authenticated_sender_force to the smtp transport.
875 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
877 PH/49 Installed latest patch from the Sieve maintainer.
879 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
880 was also set, the appendfile transport was unnecessarily scanning a
881 directory of message files (e.g. for maildir delivery) to find the count
882 of files (along with the size), even though it did not need this
883 information. It now does the scan only if it needs to find either the
884 size of the count of files.
886 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
888 PH/52 Two bugs concerned with error handling when the smtp transport is
891 (i) Exim was not creating retry information for temporary errors given
892 for individual recipients after the DATA command when the smtp transport
893 was used in LMTP mode. This meant that they could be retried too
894 frequently, and not timed out correctly.
896 (ii) Exim was setting the flag that allows error details to be returned
897 for LMTP errors on RCPT commands, but not for LMTP errors for individual
898 recipients that were returned after the DATA command.
900 PH/53 This is related to PH/52, but is more general: for any failing address,
901 when detailed error information was permitted to be returned to the
902 sender, but the error was temporary, then after the final timeout, only
903 "retry timeout exceeded" was returned. Now it returns the full error as
904 well as "retry timeout exceeded".
906 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
907 do this, and (what is worse) MTAs that accept it.
909 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
910 will now be deprecated.
912 PH/56 New os.c-cygwin from the Cygwin maintainer.
914 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
915 formats (previously only available in exim formats via -bpr, -bpru,
916 and -bpra. Now also available in native and exiqgrep formats)
918 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
919 with very large, slow to parse queues
921 JJ/08 exipick: added ! as generic prefix to negate any criteria format
923 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
925 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
926 responses to authentication challenges, though it was showing the
927 challenges; (ii) I've removed the CR characters from the debug output for
930 PH/58 Allow for the insertion of a newline as well as a space when a string
931 is turned into more than one encoded-word during RFC 2047 encoding. The
932 Sieve code now uses this.
934 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
935 data_4xx, lost_connection, tls_required.
937 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
938 message was being sent as an SMTP response.
940 PH/61 Add -l and -k options to exicyclog.
942 PH/62 When verifying, if an address was redirected to one new address, so that
943 verification continued, and the new address failed or deferred after
944 having set something in $address_data, the value of $address_data was not
945 passed back to the ACL. This was different to the case when no
946 redirection occurred. The value is now passed back in both cases.
948 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
949 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
950 use login_cap.h, so on its own it isn't the distinguishing feature. The
951 new name refers directly to the setclassresources() function.
953 PH/65 Added configuration files for NetBSD3.
955 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
957 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
958 is preferred over IPv4.
960 PH/68 The bounce_return_message and bounce_return_body options were not being
961 honoured for bounces generated during the reception of non-SMTP messages.
962 In particular, this applied to messages rejected by the ACL. This bug has
963 been fixed. However, if bounce_return_message is true and bounce_return_
964 body is false, the headers that are returned for a non-SMTP message
965 include only those that have been read before the error was detected.
966 (In the case of an ACL rejection, they have all been read.)
968 PH/69 The HTML version of the specification is now built in a directory called
969 spec_html instead of spec.html, because the latter looks like a path with
970 a MIME-type, and this confuses some software.
972 PH/70 Catch two compiler warnings in sieve.c.
974 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
975 function verify_get_ident() calls ip_connect() to connect a socket, but
976 if the "connect()" function timed out, ip_connect() used to close the
977 socket. However, verify_get_ident() also closes the socket later, and in
978 between Exim writes to the log, which may get opened at this point. When
979 the socket was closed in ip_connect(), the log could get the same file
980 descriptor number as the socket. This naturally causes chaos. The fix is
981 not to close the socket in ip_connect(); the socket should be closed by
982 the function that creates it. There was only one place in the code where
983 this was missing, in the iplookup router, which I don't think anybody now
984 uses, but I've fixed it anyway.
986 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
987 well as to direct DNS lookups. Otherwise the handling of names in host
988 lists is inconsistent and therefore confusing.
994 PH/01 Two changes to the default runtime configuration:
996 (1) Move the checks for relay_from_hosts and authenticated clients from
997 after to before the (commented out) DNS black list checks.
999 (2) Add control=submission to the relay_from_hosts and authenticated
1000 clients checks, on the grounds that messages accepted by these
1001 statements are most likely to be submissions.
1003 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1005 (1) Generate an error if the third argument for the ${prvs expansion is
1008 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1011 (3) Reset the variables that are obtained from the first argument of
1012 ${prvscheck and used in the second argument before leaving the code,
1013 because their memory is reclaimed, so using them afterwards may do
1016 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1017 one (it's much easier than Tom thought :-).
1019 (5) Because of (4), we can now allow for the use of $prvscheck_result
1020 inside the third argument.
1022 PH/03 For some reason, the default setting of PATH when running a command from
1023 a pipe transport was just "/usr/bin". I have changed it to
1026 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1027 anything to be listed in the output from -bV.
1029 PH/05 When a filter generated an autoreply, the entire To: header line was
1030 quoted in the delivery log line, like this:
1032 => >A.N.Other <ano@some.domain> <original@ddress> ...
1034 This has been changed so that it extracts the operative address. There
1035 may be more than one such address. If so, they are comma-separated, like
1038 => >ano@some.domain,ona@other.domain <original@ddress> ...
1040 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1041 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1042 not being looked up in the rDNS to get a host name, Exim was showing the
1043 IP address twice in Received: lines, even though the IP addresses were
1044 identical. For example:
1046 Received: from [1.2.3.4] (helo=[1.2.3.4])
1048 However, if the real host name was known, it was omitting the HELO data
1049 if it matched the actual IP address. This has been tidied up so that it
1050 doesn't show the same IP address twice.
1052 PH/07 When both +timestamp and +memory debugging was on, the value given by
1053 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1054 called by the debug printing, thereby overwriting the timestamp buffer.
1055 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1057 PH/08 When the original message was included in an autoreply transport, it
1058 always said "this is a copy of the message, including all the headers",
1059 even if body_only or headers_only was set. It now gives an appropriate
1062 PH/09 Applied a patch from the Sieve maintainer which:
1064 o fixes some comments
1065 o adds the (disabled) notify extension core
1066 o adds some debug output for the result of if/elsif tests
1067 o points to the current vacation draft in the documentation
1068 and documents the missing references header update
1072 o fixes a bug in processing the envelope test (when testing
1073 multiple envelope elements, the last element determinted the
1076 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1077 Electronic Mail") by including:
1079 Auto-submitted: auto-generated
1081 in the messages that it generates (bounce messages and others, such as
1082 warnings). In the case of bounce messages for non-SMTP mesages, there was
1083 also a typo: it was using "Auto_submitted" (underscore instead of
1084 hyphen). Since every message generated by Exim is necessarily in response
1085 to another message, thes have all been changed to:
1087 Auto-Submitted: auto-replied
1089 in accordance with these statements in the RFC:
1091 The auto-replied keyword:
1093 - SHOULD be used on messages sent in direct response to another
1094 message by an automatic process,
1096 - MUST NOT be used on manually-generated messages,
1098 - MAY be used on Delivery Status Notifications (DSNs) and Message
1099 Disposition Notifications (MDNs),
1101 - MUST NOT be used on messages generated by automatic or periodic
1102 processes, except for messages which are automatic responses to
1105 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1106 to the default Received: header definition.
1108 PH/12 Added log selector acl_warn_skipped (default on).
1110 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1111 variables because (a) they are in the wrong storage pool and (b) even if
1112 they were copied, it wouldn't work properly because of the caching.
1114 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1115 checking when decoding. Apparently there are clients that generate
1116 overlong encoded strings. Why am I not surprised?
1118 PH/15 If the first argument of "${if match_address" was not empty, but did not
1119 contain an "@" character, Exim crashed. Now it writes a panic log message
1120 and treats the condition as false.
1122 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1124 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1125 extension "envelope-auth". The code is finished and in agreement with
1126 other implementations, but there is no documentation so far and in fact,
1127 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1128 not changing the active code.
1130 Print executed "if" and "elsif" statements when debugging is used. This
1131 helps a great deal to understand what a filter does.
1133 Document more things not specified clearly in RFC3028. I had all this
1134 sorted out, when out of a sudden new issues came to my mind. Oops."
1136 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1139 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1140 canonical form (as documented). However, after a host name lookup from
1141 the IP address, check_host() was doing a simple string comparison with
1142 addresses acquired from the DNS when checking that the found name did
1143 have the original IP as one of its addresses. Since any found IPv6
1144 addresses are likely to be in abbreviated form, the comparison could
1145 fail. Luckily, there already exists a function for doing the comparison
1146 by converting both addresses to binary, so now that is used instead of
1147 the text comparison.
1149 PH/20 There was another similar case to PH/19, when a complete host name was
1150 given in a host list; looking up its IP address could give an abbreviated
1151 form, whereas the current host's name might or might not be abbreviated.
1152 The same fix has been applied.
1158 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1159 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1162 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1163 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1165 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1167 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1168 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1169 Also, for "Auto-Submitted", treat anything other than "no" as
1170 non-personal, in accordance with RFC 3834. (Previously it treated
1171 anything starting "auto-" as non-personal.)
1173 TF/01 The control=submission/name=... option had a problem with syntax
1174 errors if the name included a slash character. The /name= option
1175 now slurps the rest of the string, so it can include any characters
1176 but it must come last in the list of options (after /sender_retain
1179 PH/05 Some modifications to the interface to the fake nameserver for the new
1187 TK/01 Added the "success_on_redirect" address verification option. See
1188 NewStuff for rationale and an example.
1190 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1192 PH/02 Patch to exigrep to allow it to work on syslog lines.
1194 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1195 fread() to skip over the body file's header line, because in Cygwin the
1196 header line is locked and is inaccessible.
1198 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1199 co-exist for some time) to make it clear that it is the Exim ID that is
1200 referenced, not the Message-ID: header line.
1202 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1203 string_format() function, because snprintf() does not exist on all
1206 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1207 ${stat: expansion item.
1209 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1210 protocol synchronization error", to keep the pedants happy.
1212 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1213 well as for IRIX systems, when gcc is being used. See the host.c source
1216 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1218 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1221 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1222 transport and to the smtp transport in LMTP mode.
1224 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1226 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1227 run a filter in a subprocess. This could lead to confusion in subsequent
1228 lookups in the parent process. There should also be a search_tidyup() at
1229 the end of the subprocess.
1231 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1232 only if the host matched helo_try_verify_hosts, which caused the
1233 verification to occur when the EHLO/HELO command was issued. The ACL just
1234 tested the remembered result. Now, if a previous verification attempt has
1235 not happened, "verify = helo" does it there and then.
1237 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1239 TK/03 Fix log output including CR from clamd.
1241 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1242 exist provoked a memory error which could cause a segfault.
1244 PH/15 Installed PCRE 6.2
1246 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1248 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1249 of the problem. Specifically, suggested +O2 rather than +O1 for the
1252 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1254 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1255 record, the port was not correctly logged when the outgoing_port log
1256 selector was set (it logged the transort's default port).
1258 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1259 fallback_hosts, and "hosts" in the smtp transport.
1261 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1262 host errors such as "Connection refused".
1264 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1265 authentication with radiusclient 0.4.9:
1267 - Error returned from rc_read_config was caught wrongly
1268 - Username/password not passed on to radius server due to wrong length.
1270 The presumption is that some radiusclient API changes for 4.51/PH/17
1271 were not taken care of correctly. The code is still untested by me (my
1272 Linux distribution still has 0.3.2 of radiusclient), but it was
1273 contributed by a Radius user.
1275 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1276 expanding the "port" option of the smtp transport.
1278 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1279 while reading a MIME header. Thanks to Tom Hughes for a patch.
1281 PH/24 Include config.h inside local_scan.h so that configuration settings are
1284 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1285 that in an ACL for RCPT (for example), you can examine exactly what was
1288 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1289 commands, but it was not correctly comparing the address with the actual
1290 client host address. Thus, it would show the EHLO address in Received:
1291 header lines when this was not necessary.
1293 PH/27 Added the % operator to ${eval:}.
1295 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1296 it should be ignoring failures (because with -C, for example, it has lost
1297 privilege). It wasn't ignoring creation failures other than "already
1300 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1303 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1304 because an input line was too long, either on its own, or by virtue of
1305 too many continuations, the temporary file was not being removed, and the
1306 return code was incorrect.
1308 PH/31 Missing "BOOL" in function definition in filtertest.c.
1310 PH/32 Applied Sieve patches from the maintainer.
1312 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1314 PH/33 Added "verify = not_blind".
1316 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1317 Local/Makefile (with some defaults set). These are used in built scripts
1318 such as exicyclog, but they have never been used in the exim_install
1319 script (though there are many overriding facilities there). I have
1320 arranged that the exim_install script now takes note of these two
1323 PH/35 Installed configuration files for Dragonfly.
1325 PH/36 When a locally submitted message by a trusted user did not contain a
1326 From: header, and the sender address was obtained from -f or from an SMTP
1327 MAIL command, and the trusted user did not use -F to supply a sender
1328 name, $originator_name was incorrectly used when constructing a From:
1329 header. Furthermore, $originator_name was used for submission mode
1330 messages from external hosts without From: headers in a similar way,
1331 which is clearly wrong.
1333 PH/37 Added control=suppress_local_fixups.
1335 PH/38 When log_selector = +received_sender was set, and the addition of the
1336 sender made the log line's construction buffer exactly full, or one byte
1337 less than full, an overflow happened when the terminating "\n" was
1340 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1341 when the result of a list match is failure because a DNS lookup failed.
1343 PH/40 RM_COMMAND is now used in the building process.
1345 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1346 the "build-* directories that it finds.
1348 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1349 address in a domain literal was a prefix of an interface address.
1351 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1352 when verifying a sender address, unless rewrite_headers is false.
1354 PH/44 Wrote a long comment about why errors_to addresses are verified as
1355 recipients, not senders.
1357 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1358 the ratelimit ACL was added.
1360 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1362 PH/46 Added extra information about PostgreSQL errors to the error string.
1364 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1365 avoiding the need to install special zones in a real server. This is
1366 backwards compatible; if it can't find the fake resolver, it drops back.
1367 Thus, both old and new test suites can be run.
1369 TF/02 Added util/ratelimit.pl
1371 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1372 clock is set back in time.
1374 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1375 Candler <B.Candler@pobox.com>.
1377 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1378 OK for addresses that are the result of redirections.
1380 PH/48 A number of further additions for the benefit of the new test suite,
1381 including a fake gethostbyname() that interfaces to the fake DNS resolver
1384 TF/06 The fix for widen_domains has also been applied to qualify_single and
1385 search_parents which are the other dnslookup options that can cause
1388 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1391 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1393 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1394 service" if it encountered an MX record with an empty target hostname.
1395 The message is now "an MX or SRV record indicated no SMTP service".
1397 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1398 if the DNS of the sending site is misconfigured. This is quite a
1399 common situation. This change restores the behaviour of treating a
1400 helo verification defer as a failure.
1402 PH/51 If self=fail was set on a router, the bounce message did not include the
1403 actual error message.
1409 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1411 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1412 command itself ended in error, the underlying message about the transport
1413 filter timeout was being overwritten with the pipe command error. Now the
1414 underlying error message should be appended to the second error message.
1416 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1418 PH/02 Reduce the amount of output that "make" produces by default. Full output
1419 can still be requested.
1421 PH/03 The warning log line about a condition test deferring for a "warn" verb
1422 was being output only once per connection, rather than after each
1423 occurrence (because it was using the same function as for successful
1424 "warn" verbs). This seems wrong, so I have changed it.
1426 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1427 it should not have, which might have caused a crash in the right
1428 circumstances, but probably never did.
1430 PH/04 Installed a modified version of Tony Finch's patch to make submission
1431 mode fix the return path as well as the Sender: header line, and to
1432 add a /name= option so that you can make the user's friendly name appear
1435 TF/03 Added the control = fakedefer ACL modifier.
1437 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1438 Mark Lowes for thorough testing.
1440 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1442 TK/03 Merged latest SRS patch from Miles Wilton.
1444 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1445 with the definition in sysexits.h (which is #included earlier).
1446 Fortunately, Exim does not actually use EX_OK. The code used to try to
1447 preserve the sysexits.h value, by assumimg that macro definitions were
1448 scanned for macro replacements. I have been disabused of this notion,
1449 so now the code just undefines EX_OK before #including unistd.h.
1451 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1452 in the smtp transport. When a block could not be written in a single
1453 write() function, the timeout was being re-applied to each part-write.
1454 This seems wrong - if the receiver was accepting one byte at a time it
1455 would take for ever. The timeout is now adjusted when this happens. It
1456 doesn't have to be particularly precise.
1458 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1459 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1461 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1462 without a domain if the check to <postmaster@domain> fails.
1464 SC/01 Eximstats: added -xls and the ability to specify output files
1465 (patch written by Frank Heydlauf).
1467 SC/02 Eximstats: use FileHandles for outputing results.
1469 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1471 SC/04 Eximstats: fixed display of large numbers with -nvr option
1473 SC/05 Eximstats: fixed merging of reports with empty tables.
1475 SC/06 Eximstats: added the -include_original_destination flag
1477 SC/07 Eximstats: removed tabs and trailing whitespace.
1479 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1481 TK/06 MBOX spool code: Add real "From " MBOX separator line
1482 so the .eml file is really in mbox format (even though
1483 most programs do not really care). Patch from Alex Miller.
1485 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1486 The latter is generated from $received_to and is only set if the
1487 message has one envelope recipient. SA can use these headers,
1488 obviously out-of-the-box. Patch from Alex Miller.
1490 PH/08 The ${def test on a variable was returning false if the variable's
1491 value was "0", contrary to what the specification has always said!
1492 The result should be true unless the variable is empty.
1494 PH/09 The syntax error of a character other than { following "${if
1495 def:variable_name" (after optional whitespace) was not being diagnosed.
1496 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1497 accidental colon was present, for example, could give incorrect results.
1499 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1500 result is used (not including appendfile, where other changes are about
1503 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1504 This involved changing a lot of size variables from int to off_t. It
1505 should work with maildirs and everything.
1507 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1508 spamd dying while we are connected to it.
1510 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1513 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1514 function for BDB 4.3.
1516 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1518 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1519 the sources for PCRE have been re-arranged and more files are now
1522 PH/15 The code I had for printing potentially long long variables in PH/11
1523 above was not the best (it lost precision). The length of off_t variables
1524 is now inspected at build time, and an appropriate printing format (%ld
1525 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1526 to be "long long int" or "long int". This is needed for the internal
1527 formatting function string_vformat().
1529 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1530 the configuration file to be ":syslog", then the script "guesses" where
1531 the logs files are, rather than using the compiled in default. In our
1532 case the guess is not the same as the compiled default, so the script
1533 suddenly stopped working when I started to use syslog. The patch checks
1534 to see if log_file_path is "". If so, it attempts to read it from exim
1535 with no configuration file to get the compiled in version, before it
1536 falls back to the previous guessing code."
1538 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1539 implementing BATV in an Exim configuration. See NewStuff for the gory
1542 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1543 Makefile that are specific to HP-UX.
1545 PH/18 If the "use_postmaster" option was set for a recipient callout together
1546 with the "random" option, the postmaster address was used as the MAIL
1547 FROM address for the random test, but not for the subsequent recipient
1548 test. It is now used for both.
1550 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1551 patch removes a few documentation additions to RFC 3028, because the
1552 latest draft now contains them. It adds the new en;ascii-case comparator
1553 and a new error check for 8bit text in MIME parts. Comparator and
1554 require names are now matched exactly. I enabled the subaddress
1555 extension, but it is not well tested yet (read: it works for me)."
1557 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1558 rework some of the code of TK/09 above to avoid the hardwired use of
1559 "%lld" and "long long". Replaced the call to snprintf() with a call to
1562 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1563 records point to non-existent hosts", "retry timeout exceeded", and
1564 "retry time not reached for any host after a long failure period".
1566 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1567 experimental DomainKeys support:
1569 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1570 (2) On an error such as an illegally used "control", the wrong name for
1571 the control was given.
1573 These problems did NOT occur unless DomainKeys support was compiled.
1575 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1577 PH/24 Added ${if match_ip condition.
1579 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1580 databases so that it will be absolutely obvious if a crash occurs in the
1581 DB library. This is a regular occurrence (often caused by mis-matched
1584 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1585 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1586 picked up on a user's system that detects such things. There doesn't seem
1587 to be a gcc warning option for this - only an attribute that has to be
1588 put on the function's prototype. It seems that in Fedora Core 4 they have
1589 set this on a number of new functions. No doubt there will be more in due
1592 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1593 specify a transport. However, if an address that was verified by such a
1594 router was the subject of a callout, Exim crashed because it tried to
1595 read the rcpt_include_affixes from the non-existent transport. Now it
1596 just assumes that the setting of that option is false. This bug was
1597 introduced by 4.51/PH/31.
1599 PH/28 Changed -d+all to exclude +memory, because that information is very
1600 rarely of interest, but it makes the output a lot bigger. People tend to
1601 do -d+all out of habit.
1603 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1604 code in os-type was giving problems when libc.so lives in lib64, like on
1607 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1608 aren't the modern standard, and it seems that some systems' include files
1609 don't always have them. Exim was already checking for some of the newer
1610 ones like T_AAAA, and defining it itself. I've added checks for all the
1611 record types that Exim uses.
1613 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1614 not automatically generating a new one, as it is supposed to. This
1615 prevented TLS from working. If the file did exist, but contained invalid
1616 data, a new version was generated, as expected. It was only the case of a
1617 non-existent file that was broken.
1619 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1620 with a change in libdomainkeys > 0.64.
1622 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1623 from DNS. If the selector record carries the flag, it now has
1624 precedence over the domain-wide flag.
1626 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1628 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1629 the use of a port name that isn't defined in /etc/services) occurred, the
1630 message was deferred as in a normal delivery, and thus remained on the
1631 spool, instead of being failed because of the mua_wrapper setting. This
1632 is now fixed, and I tidied up some of the mua_wrapper messages at the
1635 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1636 the messages in a hash of arrays rather than using individual hashes.
1637 This is a bit cleaner and results in dramatic memory savings, albeit
1638 at a slight CPU cost.
1640 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1641 as requested by Marc Sherman.
1643 SC/10 Eximstats: added histograms for user specified patterns as requested
1646 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1648 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1649 fopen() in the content-scanning modules that did not already have it.
1655 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1656 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1658 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1660 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1662 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1663 to test Sieve filters that use "vacation".
1665 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1666 that changes the way the GnuTLS parameters are stored in the cache file.
1667 The new format can be generated externally. For backward compatibility,
1668 if the data in the cache doesn't make sense, Exim assumes it has read an
1669 old-format file, and it generates new data and writes a new file. This
1670 means that you can't go back to an older release without removing the
1673 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1674 work if there are any delivery delays because "one_time" forces the
1675 parent to be marked "delivered", so its unseen clone is never tried
1676 again. For this reason, Exim now forbids the simultaneous setting of
1679 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1680 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1681 entry if you want to know the details. The fix, however, neglected to
1682 consider the case when local delivery batching is involved. The test for
1683 "previously delivered" was not happening when checking to see if an
1684 address could be batched with a previous (undelivered) one; under
1685 certain circumstances this could lead to multiple deliveries to the same
1688 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1689 in its include files, and this causes problems building Exim.
1691 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1692 header_syntax) but Exim was just ignoring anything given after a slash.
1693 In particular, this caused confusion with an attempt to use "verify =
1694 reverse_host_lookup/defer_ok". An error is now given when options are
1695 supplied for verify items that do not have them. (Maybe reverse_host_
1696 lookup should have a defer_ok option, but that's a different point.)
1698 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1699 defined by RFC 821) to 2048, because there were problems with some AUTH
1700 commands, and RFC 1869 says the size should be increased for extended
1701 SMTP commands that take arguments.
1703 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1706 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1707 "unknown" error; now it says that the functionality isn't in the binary.
1709 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1710 an address' error message when a string expansion fails (syntax or
1711 whatever). Otherwise the password may appear in the log. Following change
1712 PH/42 below, there is no longer a chance of it appearing in a bounce
1715 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1717 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1718 of its addresses was ever tried. (Bugzilla bug #2.)
1720 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1721 the result incorrectly in the debug output. (It correctly added a newline
1722 to what was transported.)
1724 TF/01 Added $received_time.
1726 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1727 commented out examples of how to interface to a virus scanner and to
1728 SpamAssassin. Also added commented examples of av_scanner and
1729 spamd_address settings.
1731 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1732 and controls are allowed in which ACLs. There were a couple of minor
1733 errors. Some of the entries in the conditions table (which is a table of
1734 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1735 negation of where the condition IS allowed.
1737 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1739 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1740 header file does not have a version number, so I've had to invent a new
1741 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1742 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1743 radiusclient), but it was contributed by a Radius user.
1745 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1746 files or directories, for OpenSSL.
1748 PH/19 When an Exim process that is running as root has to create an Exim log
1749 file, it does so in a subprocess that runs as exim:exim so as to get the
1750 ownership right at creation (otherwise, other Exim processes might see
1751 the file with the wrong ownership). There was no test for failure of this
1752 fork() call, which would lead to the process getting stuck as it waited
1753 for a non-existent subprocess. Forks do occasionally fail when resources
1754 run out. I reviewed all the other calls to fork(); they all seem to check
1757 PH/20 When checking for unexpected SMTP input at connect time (before writing
1758 the banner), Exim was not dealing correctly with a non-positive return
1759 from the read() function. If the client had disconnected by this time,
1760 the result was a log entry for a synchronization error with an empty
1761 string after "input=" when read() returned zero. If read() returned -1
1762 (an event I could not check), uninitialized data bytes were printed.
1763 There were reports of junk text (parts of files, etc) appearing after
1766 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1768 PH/22 Added support for macro redefinition, and (re)definition in between
1769 driver and ACL definitions.
1771 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1772 forgetting to use the resulting value; it was using the unexpanded value.
1774 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1775 hadn't been configured. The fix is from Juergen Kreileder, who
1776 understands it better than I do:
1778 "Here's what I see happening with three configured cyrus_sasl
1779 authenticators configured (plain, login, cram-md5):
1781 On startup auth_cyrus_sasl_init() gets called for each of these.
1782 This means three calls to sasl_listmech() without a specified mech_list.
1783 => SASL tests which mechs of all available mechs actually work
1784 => three warnings about OTP not working
1785 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1787 With the patch, sasl_listmech() also gets called three times. But now
1788 SASL's mech_list option is set to the server_mech specified in the the
1789 authenticator. Or in other words, the answer from sasl_listmech()
1790 gets limited to just the mech you're testing for (which is different
1792 => the return list contains just 'plain' or 'login', 'cram-md5' or
1793 nothing depending on the value of ob->server_mech.
1795 I've just tested the patch: Authentication still works fine,
1796 unavailable mechs specified in the exim configuration are still
1797 caught, and the auth.log warnings about OTP are gone."
1799 PH/25 When debugging is enabled, the contents of the command line are added
1800 to the debugging output, even when log_selector=+arguments is not
1803 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1804 answer is "GNU", and only if the return is "GNU/something" is the answer
1807 PH/27 $acl_verify_message is now set immediately after the failure of a
1808 verification in an ACL, and so is available in subsequent modifiers. In
1809 particular, the message can be preserved by coding like this:
1811 warn !verify = sender
1812 set acl_m0 = $acl_verify_message
1814 Previously, $acl_verify_message was set only while expanding "message"
1815 and "log_message" when a very denied access.
1817 PH/28 Modified OS/os.c-Linux with
1819 -#ifndef OS_LOAD_AVERAGE
1820 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1822 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1823 nomenclature these days.)
1825 PH/29 Installed patch from the Sieve maintainer that adds the options
1826 sieve_useraddress and sieve_subaddress to the redirect router.
1828 PH/30 In these circumstances:
1829 . Two addresses routed to the same list of hosts;
1830 . First host does not offer TLS;
1831 . First host accepts first address;
1832 . First host gives temporary error to second address;
1833 . Second host offers TLS and a TLS session is established;
1834 . Second host accepts second address.
1835 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1836 and peerdn, if requested) that were in fact used only for the second
1839 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1840 attention to any local part prefix or suffix that was matched by the
1841 router that accepted the address. It now behaves in the same way as it
1842 does for delivery: the affixes are removed from the local part unless
1843 rcpt_include_affixes is set on the transport.
1845 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1846 timeout during the DATA phase of an incoming message.
1848 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1849 applied a patch sanctioned by the Sieve maintainer.
1851 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1852 the uid or gid is negative. A case of a negative gid caused this to be
1853 noticed. The fix allows for either to be negative.
1855 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1856 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1859 PH/36 The change PH/12 above was broken. Fixed it.
1861 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1862 the grounds that routing the same address twice would always produce the
1863 same answer. This might have been true once, but it is certainly no
1864 longer true now. Routing a child address may depend on the previous
1865 routing that produced that child. Some complicated redirection strategies
1866 went wrong when messages had multiple recipients, and made Exim's
1867 behaviour dependent on the order in which the addresses were given.
1869 I have moved the duplicate checking until after the routing is complete.
1870 Exim scans the addresses that are assigned to local and remote
1871 transports, and removes any duplicates. This means that more work will be
1872 done, as duplicates will always all be routed, but duplicates are
1873 presumably rare, so I don't expect this is of any significance.
1875 For deliveries to pipes, files, and autoreplies, the duplicate checking
1876 still happens during the routing process, since they are not going to be
1879 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1880 It corrects a timeout issue with spamd. This is Ian's comment: "The
1881 background is that sometimes spamd either never reads data from a
1882 connection it has accepted, or it never writes response data. The exiscan
1883 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1884 blindly assumes that writes won't block so it may never time out."
1886 PH/39 Allow G after quota size as well as K and M.
1888 PH/40 The value set for $authenticated_id in an authenticator may not contain
1889 binary zeroes or newlines because the value is written to log lines and
1890 to spool files. There was no check on this. Now the value is run through
1891 the string_printing() function so that such characters are converted to
1892 printable escape sequences.
1894 PH/41 $message_linecount is a new variable that contains the total number of
1895 lines in the message. Compare $body_linecount, which is the count for the
1898 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1899 bounce and delay warning messages, except in certain special cases, which
1902 (a) An SMTP error message from a remote host;
1903 (b) A message specified in a :fail: redirection;
1904 (c) A message specified in a "fail" command in a system filter;
1905 (d) A message specified in a FAIL return from the queryprogram router;
1906 (e) A message specified by the cannot_route_message router option.
1908 In these cases only, Exim does include the error details in bounce and
1909 warning messages. There are also a few cases where bland messages such
1910 as "unrouteable address" or "local delivery error" are given.
1912 PH/43 $value is now also set for the "else" part of a ${run expansion.
1914 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1915 being worked on, but at least Exim now implements the latest version to
1918 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1919 process to complete was treated as a delivery failure, a timeout while
1920 writing the message to the pipe was logged, but erroneously treated as a
1921 successful delivery. Such timeouts include transport filter timeouts. For
1922 consistency with the overall process timeout, these timeouts are now
1923 treated as errors, giving rise to delivery failures by default. However,
1924 there is now a new Boolean option for the pipe transport called
1925 timeout_defer, which, if set TRUE, converts the failures into defers for
1926 both kinds of timeout. A transport filter timeout is now identified in
1929 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1930 systems where "make" and "gmake" are different, calling "gmake" at top
1931 level broke things. I've arranged for the value of $(MAKE) to be passed
1932 from the Makefile to this script so that it can call the same version of
1936 A note about Exim versions 4.44 and 4.50
1937 ----------------------------------------
1939 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1940 changes of various kinds. As a consequence, a big documentation update was
1941 needed. This delayed the release for rather longer than seemed good, especially
1942 in the light of a couple of (minor) security issues. Therefore, the changes
1943 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1944 release. So 4.44 and 4.50 are in effect two different branches that both start
1947 I have left the 4.50 change log unchanged; it contains all the changes since
1948 4.43. The change log for 4.44 is below; many of its items are identical to
1949 those for 4.50. This seems to be the most sensible way to preserve the
1950 historical information.
1956 1. Minor wording change to the doc/README.SIEVE file.
1958 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1959 computation of the current number of files was incorrect.
1961 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1962 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1965 4. Give more explanation in the error message when the command for a transport
1966 filter fails to execute.
1968 5. There are several places where Exim runs a non-Exim command in a
1969 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1970 was being done only for the command run by the queryprogram router. It is
1971 now done for all such subprocesses. The other cases are: ${run, transport
1972 filters, and the commands run by the lmtp and pipe transports.
1974 6. Added CONFIGURE_GROUP build-time option.
1976 7. Some older OS have a limit of 256 on the maximum number of file
1977 descriptors. Exim was using setrlimit() to set 1000 as a large value
1978 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1979 systems. I've change it so that if it can't get 1000, it tries for 256.
1981 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1982 was an oversight, and furthermore, ever since the addition of extra
1983 controls (e.g. 4.43/32), the checks on when to allow different forms of
1984 "control" were broken. There should now be diagnostics for all cases when a
1985 control that does not make sense is encountered.
1987 9. Added the /retain_sender option to "control=submission".
1989 10. $recipients is now available in the predata ACL (oversight).
1991 11. Tidy the search cache before the fork to do a delivery from a message
1992 received from the command line. Otherwise the child will trigger a lookup
1993 failure and thereby defer the delivery if it tries to use (for example) a
1994 cached ldap connection that the parent has called unbind on.
1996 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1997 of $address_data from the recipient verification was clobbered by the
1998 sender verification.
2000 13. The value of address_data from a sender verification is now available in
2001 $sender_address_data in subsequent conditions in the ACL statement.
2003 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2005 15. Added a new option "connect=<time>" to callout options, to set a different
2008 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2009 was its contents. (It was OK if the option was not defined at all.)
2011 17. A "Completed" log line is now written for messages that are removed from
2012 the spool by the -Mrm option.
2014 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2015 information about exactly what failed.
2017 19. Added -dd to debug only the daemon process.
2019 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2020 handles timeouts, both on the server side and network timeouts. Renamed the
2021 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2023 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2024 It is now set to "smtps".
2026 22. $host_address is now set to the target address during the checking of
2027 ignore_target_hosts.
2029 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2030 being passed; this would have caused $sender_host_name to have been used if
2031 matching the list had actually called for a host name (not very likely,
2032 since this list is usually IP addresses). A host name is now passed as
2035 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2036 code that specifies a non-restarting handler (typically sigaction() in
2037 modern systems) in an attempt to fix a rare and obscure crash bug.
2039 25. Narrowed the window for a race in the daemon that could cause it to ignore
2040 SIGCHLD signals. This is not a major problem, because they are used only to
2041 wake it up if nothing else does.
2043 26. A malformed maildirsize file could cause Exim to calculate negative values
2044 for the mailbox size or file count. Odd effects could occur as a result.
2045 The maildirsize information is now recalculated if the size or filecount
2048 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2049 support for a long time. Removed HAVE_SYS_VFS_H.
2051 28. Installed the latest version of exipick from John Jetmore.
2053 29. In an address list, if the pattern was not a regular expression, an empty
2054 subject address (from a bounce message) matched only if the pattern was an
2055 empty string. Non-empty patterns were not even tested. This was the wrong
2056 because it is perfectly reasonable to use an empty address as part of a
2057 database query. An empty address is now tested by patterns that are
2058 lookups. However, all the other forms of pattern expect the subject to
2059 contain a local part and a domain, and therefore, for them, an empty
2060 address still always fails if the pattern is not itself empty.
2062 30. Exim went into a mad DNS loop when attempting to do a callout where the
2063 host was specified on an smtp transport, and looking it up yielded more
2064 than one IP address.
2066 31. Re-factored the code for checking spool and log partition space into a
2067 function that finds that data and another that does the check. The former
2068 is then used to implement four new variables: $spool_space, $log_space,
2069 $spool_inodes, and $log_inodes.
2071 32. The RFC2047 encoding function was originally intended for short strings
2072 such as real names; it was not keeping to the 75-character limit for
2073 encoded words that the RFC imposes. It now respects the limit, and
2074 generates multiple encoded words if necessary. To be on the safe side, I
2075 have increased the buffer size for the ${rfc2047: expansion operator from
2078 33. It is now permitted to omit both strings after an "if" condition; if the
2079 condition is true, the result is "true". As before, when the second string
2080 is omitted, a false condition yields an empty string. This makes it less
2081 cumbersome to write custom ACL and router conditions.
2083 34. Failure to deliver a bounce message always caused it to be frozen, even if
2084 there was an errors_to setting on the router. The errors_to setting is now
2087 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2088 canonical form (fully expanded) before being placed in
2089 $sender_host_address.
2091 36. The table in the code that translates DNS record types into text (T_A to
2092 "A" for instance) was missing entries for NS and CNAME. It is just possible
2093 that this could have caused confusion if both these types were looked up
2094 for the same domain, because the text type is used as part of Exim's
2095 per-process caching. But the chance of anyone hitting this buglet seems
2098 37. The dnsdb lookup has been extended in a number of ways.
2100 (1) There is a new type, "zns", which walks up the domain tree until it
2101 finds some nameserver records. It should be used with care.
2103 (2) There is a new type, "mxh", which is like "mx" except that it returns
2104 just the host names, not the priorities.
2106 (3) It is now possible to give a list of domains (or IP addresses) to be
2107 looked up. The behaviour when one of the lookups defers can be
2108 controlled by a keyword.
2110 (4) It is now possible to specify the separator character for use when
2111 multiple records are returned.
2113 38. The dnslists ACL condition has been extended: it is now possible to supply
2114 a list of IP addresses and/or domains to be looked up in a particular DNS
2117 39. Added log_selector=+queue_time_overall.
2119 40. When running the queue in the test harness, wait just a tad after forking a
2120 delivery process, to get repeatability of debugging output.
2122 41. Include certificate and key file names in error message when GnuTLS fails
2123 to set them up, because the GnuTLS error message doesn't include the name
2124 of the failing file when there is a problem reading it.
2126 42. Allow both -bf and -bF in the same test run.
2128 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2130 44. The "Exiscan patch" is now merged into the mainline Exim source.
2132 45. Sometimes the final signoff response after QUIT could fail to get
2133 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2134 < 0 before doing a fflush(). This bug looks as though it goes back to the
2135 introduction of TLS in release 3.20, but "sometimes" must have been rare
2136 because the tests only now provoked it.
2138 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2139 (this can affect the format of dates).
2141 47. exim_tidydb, when checking for the continued existence of a message for
2142 which it has found a message-specific retry record, was not finding
2143 messages that were in split spool directories. Consequently, it was
2144 deleting retry records that should have stayed in existence.
2146 48. Steve fixed some bugs in eximstats.
2148 49. The SPA authentication driver was not abandoning authentication and moving
2149 on to the next authenticator when an expansion was forced to fail,
2150 contradicting the general specification for all authenticators. Instead it
2151 was generating a temporary error. It now behaves as specified.
2153 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2154 (the order specifies the preference for clients). The order is now AES256,
2155 AES128, 3DES, ARCFOUR128.
2157 51. Small patch to Sieve code - explicitly set From: when generating an
2160 52. Exim crashed if a remote delivery caused a very long error message to be
2161 recorded - for instance if somebody sent an entire SpamAssassin report back
2162 as a large number of 550 error lines. This bug was coincidentally fixed by
2163 increasing the size of one of Exim's internal buffers (big_buffer) that
2164 happened as part of the Exiscan merge. However, to be on the safe side, I
2165 have made the code more robust (and fixed the comments that describe what
2168 53. Now that there can be additional text after "Completed" in log lines (if
2169 the queue_time_overall log selector is set), a one-byte patch to exigrep
2170 was needed to allow it to recognize "Completed" as not the last thing in
2173 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2174 patch that reportedly fixes this has been added. I am not expert enough to
2175 create a test for it. This is what the patch creator wrote:
2177 "I found a little strange behaviour of ldap code when working with
2178 Windows 2003 AD Domain, where users was placed in more than one
2179 Organization Units. When I tried to give exim partial DN, the exit code
2180 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2181 But simultaneously result of request was absolutely normal ldap result,
2182 so I produce this patch..."
2184 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2185 REFERENCE, so I have modified the code to exclude the patch when that macro
2188 55. Some experimental protocols are using DNS PTR records for new purposes. The
2189 keys for these records are domain names, not reversed IP addresses. The
2190 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2191 leaves it alone. Component reversal etc. now happens only for IP addresses.
2194 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2196 57. Double the size of the debug message buffer (to 2048) so that more of very
2197 long debug lines gets shown.
2199 58. The exicyclog utility now does better if the number of log files to keep
2200 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2202 59. Two changes related to the smtp_active_hostname option:
2204 (1) $smtp_active_hostname is now available as a variable.
2205 (2) The default for smtp_banner uses $smtp_active_hostname instead
2206 of $primary_hostname.
2208 60. The host_aton() function is supposed to be passed a string that is known
2209 to be a valid IP address. However, in the case of IPv6 addresses, it was
2210 not checking this. This is a hostage to fortune. Exim now panics and dies
2211 if the condition is not met. A case was found where this could be provoked
2212 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2213 components; fortuitously, this particular loophole had already been fixed
2214 by change 4.50/55 above.
2216 If there are any other similar loopholes, the new check in host_aton()
2217 itself should stop them being exploited. The report I received stated that
2218 data on the command line could provoke the exploit when Exim was running as
2219 exim, but did not say which command line option was involved. All I could
2220 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2221 running as the user.
2224 61. There was a buffer overflow vulnerability in the SPA authentication code
2225 (which came originally from the Samba project). I have added a test to the
2226 spa_base64_to_bits() function which I hope fixes it.
2229 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2230 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2232 63. The daemon start-up calls getloadavg() while still root for those OS that
2233 need the first call to be done as root, but it missed one case: when
2234 deliver_queue_load_max is set with deliver_drop_privilege. This is
2235 necessary for the benefit of the queue runner, because there is no re-exec
2236 when deliver_drop_privilege is set.
2238 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2239 This has been fixed.
2241 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2242 was in use, was not putting the data itself into the right store pool;
2243 consequently, it could be overwritten for a subsequent message in the same
2244 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2247 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2249 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2250 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2251 the function were treating the return as a boolean value, which happened to
2252 work because 0=false and not-0=true, but is not correct code.
2254 68. The host_aton() function was not handling scoped IPv6 addresses (those
2255 with, for example, "%eth0" on the end) correctly.
2257 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2258 negated items (that is, ~something) in unsigned ints. Some compilers
2259 apparently mutter when there is no cast.
2261 70. If an address verification called from an ACL failed, and did not produce a
2262 user-specific message (i.e. there was only a "system" message), nothing was
2263 put in $acl_verify_message. In this situation, it now puts the system
2266 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2267 session; change 4.31/43 added the unwanted input to the log line - except
2268 that it did not do this in the start of session case. It now does.
2270 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2271 This is wrong and can cause the other end to generate a synchronization
2272 error if it is another Exim or anything else that does the synchronization
2273 check. A QUIT command is no longer sent after a timeout.
2275 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2276 during host lookups.
2278 74. The defer_ok option of callout verification was not working if it was used
2279 when verifying addresses in header lines, that is, for this case:
2281 verify = header_sender/callout=defer_ok
2283 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2284 those file descriptors could be used for SMTP connections. If anything
2285 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2286 could be sent to the SMTP client, causing chaos. The daemon now opens
2287 stdin, stdout, and stderr to /dev/null when it puts itself into the
2290 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2291 log by default. The user can override this with suitable Perl magic.
2293 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2294 the log message when discard triggers, was not working for the DATA ACL or
2295 for the non-SMTP ACL.
2297 78. Error message wording change in sieve.c.
2299 79. If smtp_accept_max_per_host was set, the number of connections could be
2300 restricted to fewer than expected, because the daemon was trying to set up
2301 a new connection before checking whether the processes handling previous
2302 connections had finished. The check for completed processes is now done
2303 earlier. On busy systems, this bug wouldn't be noticed because something
2304 else would have woken the daemon, and it would have reaped the completed
2307 80. If a message was submitted locally by a user whose login name contained one
2308 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2309 It caused a spool format error. I have fixed the spool reading code. A
2310 related problem was that the "from" clause in the Received: line became
2311 illegal because of the space(s). It is now covered by ${quote_local_part.
2313 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2316 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2318 83. Patch from Sieve maintainer to fix unterminated string problem in
2319 "vacation" handling.
2321 84. Some minor changes to the Linux configuration files to help with other
2322 OS variants using glibc.
2324 85. One more patch for Sieve to update vacation handling to latest spec.
2327 ----------------------------------------------------
2328 See the note above about the 4.44 and 4.50 releases.
2329 ----------------------------------------------------
2335 1. Change 4.43/35 introduced a bug that caused file counts to be
2336 incorrectly computed when quota_filecount was set in an appendfile
2339 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2340 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2343 3. Give more explanation in the error message when the command for a transport
2344 filter fails to execute.
2346 4. There are several places where Exim runs a non-Exim command in a
2347 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2348 was being done only for the command run by the queryprogram router. It is
2349 now done for all such subprocesses. The other cases are: ${run, transport
2350 filters, and the commands run by the lmtp and pipe transports.
2352 5. Some older OS have a limit of 256 on the maximum number of file
2353 descriptors. Exim was using setrlimit() to set 1000 as a large value
2354 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2355 systems. I've change it so that if it can't get 1000, it tries for 256.
2357 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2358 was an oversight, and furthermore, ever since the addition of extra
2359 controls (e.g. 4.43/32), the checks on when to allow different forms of
2360 "control" were broken. There should now be diagnostics for all cases when a
2361 control that does not make sense is encountered.
2363 7. $recipients is now available in the predata ACL (oversight).
2365 8. Tidy the search cache before the fork to do a delivery from a message
2366 received from the command line. Otherwise the child will trigger a lookup
2367 failure and thereby defer the delivery if it tries to use (for example) a
2368 cached ldap connection that the parent has called unbind on.
2370 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2371 of $address_data from the recipient verification was clobbered by the
2372 sender verification.
2374 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2375 was its contents. (It was OK if the option was not defined at all.)
2377 11. A "Completed" log line is now written for messages that are removed from
2378 the spool by the -Mrm option.
2380 12. $host_address is now set to the target address during the checking of
2381 ignore_target_hosts.
2383 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2384 being passed; this would have caused $sender_host_name to have been used if
2385 matching the list had actually called for a host name (not very likely,
2386 since this list is usually IP addresses). A host name is now passed as
2389 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2390 code that specifies a non-restarting handler (typically sigaction() in
2391 modern systems) in an attempt to fix a rare and obscure crash bug.
2393 15. Narrowed the window for a race in the daemon that could cause it to ignore
2394 SIGCHLD signals. This is not a major problem, because they are used only to
2395 wake it up if nothing else does.
2397 16. A malformed maildirsize file could cause Exim to calculate negative values
2398 for the mailbox size or file count. Odd effects could occur as a result.
2399 The maildirsize information is now recalculated if the size or filecount
2402 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2403 support for a long time. Removed HAVE_SYS_VFS_H.
2405 18. Updated exipick to current release from John Jetmore.
2407 19. Allow an empty sender to be matched against a lookup in an address list.
2408 Previously the only cases considered were a regular expression, or an
2411 20. Exim went into a mad DNS lookup loop when doing a callout where the
2412 host was specified on the transport, if the DNS lookup yielded more than
2415 21. The RFC2047 encoding function was originally intended for short strings
2416 such as real names; it was not keeping to the 75-character limit for
2417 encoded words that the RFC imposes. It now respects the limit, and
2418 generates multiple encoded words if necessary. To be on the safe side, I
2419 have increased the buffer size for the ${rfc2047: expansion operator from
2422 22. Failure to deliver a bounce message always caused it to be frozen, even if
2423 there was an errors_to setting on the router. The errors_to setting is now
2426 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2427 canonical form (fully expanded) before being placed in
2428 $sender_host_address.
2430 24. Updated eximstats to version 1.33
2432 25. Include certificate and key file names in error message when GnuTLS fails
2433 to set them up, because the GnuTLS error message doesn't include the name
2434 of the failing file when there is a problem reading it.
2436 26. Expand error message when OpenSSL has problems setting up cert/key files.
2439 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2440 (this can affect the format of dates).
2442 28. exim_tidydb, when checking for the continued existence of a message for
2443 which it has found a message-specific retry record, was not finding
2444 messages that were in split spool directories. Consequently, it was
2445 deleting retry records that should have stayed in existence.
2447 29. eximstats updated to version 1.35
2448 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2449 1.35 - bugfix such that pie charts by volume are generated correctly
2451 30. The SPA authentication driver was not abandoning authentication and moving
2452 on to the next authenticator when an expansion was forced to fail,
2453 contradicting the general specification for all authenticators. Instead it
2454 was generating a temporary error. It now behaves as specified.
2456 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2457 (the order specifies the preference for clients). The order is now AES256,
2458 AES128, 3DES, ARCFOUR128.
2460 31. Small patch to Sieve code - explicitly set From: when generating an
2463 32. Exim crashed if a remote delivery caused a very long error message to be
2464 recorded - for instance if somebody sent an entire SpamAssassin report back
2465 as a large number of 550 error lines. This bug was coincidentally fixed by
2466 increasing the size of one of Exim's internal buffers (big_buffer) that
2467 happened as part of the Exiscan merge. However, to be on the safe side, I
2468 have made the code more robust (and fixed the comments that describe what
2471 33. Some experimental protocols are using DNS PTR records for new purposes. The
2472 keys for these records are domain names, not reversed IP addresses. The
2473 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2474 leaves it alone. Component reversal etc. now happens only for IP addresses.
2477 34. The host_aton() function is supposed to be passed a string that is known
2478 to be a valid IP address. However, in the case of IPv6 addresses, it was
2479 not checking this. This is a hostage to fortune. Exim now panics and dies
2480 if the condition is not met. A case was found where this could be provoked
2481 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2482 components; fortuitously, this particular loophole had already been fixed
2483 by change 4.50/55 or 4.44/33 above.
2485 If there are any other similar loopholes, the new check in host_aton()
2486 itself should stop them being exploited. The report I received stated that
2487 data on the command line could provoke the exploit when Exim was running as
2488 exim, but did not say which command line option was involved. All I could
2489 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2490 running as the user.
2493 35. There was a buffer overflow vulnerability in the SPA authentication code
2494 (which came originally from the Samba project). I have added a test to the
2495 spa_base64_to_bits() function which I hope fixes it.
2498 36. The daemon start-up calls getloadavg() while still root for those OS that
2499 need the first call to be done as root, but it missed one case: when
2500 deliver_queue_load_max is set with deliver_drop_privilege. This is
2501 necessary for the benefit of the queue runner, because there is no re-exec
2502 when deliver_drop_privilege is set.
2504 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2505 was in use, was not putting the data itself into the right store pool;
2506 consequently, it could be overwritten for a subsequent message in the same
2507 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2510 38. Sometimes the final signoff response after QUIT could fail to get
2511 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2512 < 0 before doing a fflush(). This bug looks as though it goes back to the
2513 introduction of TLS in release 3.20, but "sometimes" must have been rare
2514 because the tests only now provoked it.
2520 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2521 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2522 Now it returns an integer. A number of calls were still expecting a T/F
2523 return. Fortuitously, in all cases, the tests worked in OK situations,
2524 which is the norm. However, things would have gone wrong on any write
2525 failures on the smtp file descriptor. This function is used when sending
2526 messages over SMTP and also when doing verify callouts.
2528 2. When Exim is called to do synchronous delivery of a locally submitted
2529 message (the -odf or -odi options), it no longer closes stderr before doing
2532 3. Implemented the mua_wrapper option.
2534 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2536 5. Implemented the functions header_remove(), header_testname(),
2537 header_add_at_position(), and receive_remove_recipient(), and exported them
2540 6. If an ACL "warn" statement specified the addition of headers, Exim already
2541 inserted X-ACL-Warn: at the start if there was no header name. However, it
2542 was not making this test for the second and subsequent header lines if
2543 there were newlines in the string. This meant that an invalid header could
2544 be inserted if Exim was badly configured.
2546 7. Allow an ACL "warn" statement to add header lines at the start or after all
2547 the Received: headers, as well as at the end.
2549 8. Added the rcpt_4xx retry error code.
2551 9. Added postmaster_mailfrom=xxx to callout verification option.
2553 10. Added mailfrom=xxxx to the callout verification option, for verify=
2556 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2557 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2559 12. Inserted some casts to stop certain compilers warning when using pointer
2560 differences as field lengths or precisions in printf-type calls (mostly
2561 affecting debugging statements).
2563 13. Added optional readline() support for -be (dynamically loaded).
2565 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2566 same clock tick as a message's arrival, so that its received time was the
2567 same as the "first fail" time on the retry record, and that message
2568 remained on the queue past the ultimate address timeout, every queue runner
2569 would try a delivery (because it was past the ultimate address timeout) but
2570 after another failure, the ultimate address timeout, which should have then
2571 bounced the address, did not kick in. This was a "< instead of <=" error;
2572 in most cases the first failure would have been in the next clock tick
2573 after the received time, and all would be well.
2575 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2576 being recognized when the domain list was tested by the match_domain
2577 condition in an expansion string.
2579 16. Added the ${str2b64: operator.
2581 17. Exim was always calling setrlimit() to set a large limit for the number of
2582 processes, without checking whether the existing limit was already
2583 adequate. (It did check for the limit on file descriptors.) Furthermore,
2584 errors from getrlimit() and setrlimit() were being ignored. Now they are
2585 logged to the main and panic logs, but Exim does carry on, to try to do its
2586 job under whatever limits there are.
2588 18. Imported PCRE 5.0.
2590 19. Trivial typo in log message " temporarily refused connection" (the leading
2593 20. If the log selector return_path_on_delivery was set and an address was
2594 redirected to /dev/null, the delivery process crashed because it assumed
2595 that a return path would always be set for a "successful" delivery. In this
2596 case, the whole delivery is bypassed as an optimization, and therefore no
2599 21. Internal re-arrangement: the function for sending a challenge and reading
2600 a response while authentication was assuming a zero-terminated challenge
2601 string. It's now changed to take a pointer and a length, to allow for
2602 binary data in such strings.
2604 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2606 23. Exim was not respecting finduser_retries when seeking the login of the
2607 uid under which it was called; it was always trying 10 times. (The default
2608 setting of finduser_retries is zero.) Also, it was sleeping after the final
2609 failure, which is pointless.
2611 24. Implemented tls_on_connect_ports.
2613 25. Implemented acl_smtp_predata.
2615 26. If the domain in control=submission is set empty, Exim assumes that the
2616 authenticated id is a complete email address when it generates From: or
2617 Sender: header lines.
2619 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2620 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2621 chown and chgrp in /bin and hostname in /usr/bin.
2623 28. Exim was keeping the "process log" file open after each use, just as it
2624 does for the main log. This opens the possibility of it remaining open for
2625 long periods when the USR1 signal hits a daemon. Occasional processlog
2626 errors were reported, that could have been caused by this. Anyway, it seems
2627 much more sensible not to leave this file open at all, so that is what now
2630 29. The long-running daemon process does not normally write to the log once it
2631 has entered its main loop, and it closes the log before doing so. This is
2632 so that log files can straightforwardly be renamed and moved. However,
2633 there are a couple of unusual error situations where the daemon does write
2634 log entries, and I had neglected to close the log afterwards.
2636 30. The text of an SMTP error response that was received during a remote
2637 delivery was being truncated at 512 bytes. This is too short for some of
2638 the long messages that one sometimes sees. I've increased the limit to
2641 31. It is now possible to make retry rules that apply only when a message has a
2642 specific sender, in particular, an empty sender.
2644 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2645 it possible to be selective about when SMTP synchronization is enforced.
2647 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2649 32. Implemented hosts_connection_nolog.
2651 33. Added an ACL for QUIT.
2653 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2656 35. Added mailbox_size and mailbox_filecount to appendfile.
2658 36. Added control = no_multiline_responses to ACLs.
2660 37. There was a bug in the logic of the code that waits for the clock to tick
2661 in the case where the clock went backwards by a substantial amount such
2662 that the microsecond fraction of "now" was more than the microsecond
2663 fraction of "then" (but the whole seconds number was less).
2665 38. Added support for the libradius Radius client library this is found on
2666 FreeBSD (previously only the radiusclient library was supported).
2672 1. When certain lookups returned multiple values in the form name=value, the
2673 quoting of the values was not always being done properly. Specifically:
2674 (a) If the value started with a double quote, but contained no whitespace,
2676 (b) If the value contained whitespace other than a space character (i.e.
2677 tabs or newlines or carriage returns) it was not quoted.
2678 This fix has been applied to the mysql and pgsql lookups by writing a
2679 separate quoting function and calling it from the lookup code. The fix
2680 should probably also be applied to nisplus, ibase and oracle lookups, but
2681 since I cannot test any of those, I have not disturbed their existing code.
2683 2. A hit in the callout cache for a specific address caused a log line with no
2684 reason for rejecting RCPT. Now it says "Previous (cached) callout
2685 verification failure".
2687 3. There was an off-by-one bug in the queryprogram router. An over-long
2688 return line was truncated at 256 instead of 255 characters, thereby
2689 overflowing its buffer with the terminating zero. As well as fixing this, I
2690 have increased the buffer size to 1024 (and made a note to document this).
2692 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2693 when Exim is waiting for an SMTP response from a remote server, Exim
2694 restarts its select() call on the socket, thereby resetting its timeout.
2695 This is not a problem when such interrupts are rare. Somebody set up a cron
2696 job to run exiwhat every 2 minutes, which is less than the normal select()
2697 timeout (5 or 10 minutes). This meant that the select() timeout never
2698 kicked in because it was always reset. I have fixed this by comparing the
2699 time when an interrupt arrives with the time at the start of the first call
2700 to select(). If more time than the timeout has elapsed, the interrupt is
2701 treated as a timeout.
2703 5. Some internal re-factoring in preparation for the addition of Sieve
2704 extensions (by MH). In particular, the "personal" test is moved to a
2705 separate function, and given an option for scanning Cc: and Bcc: (which is
2706 not set for Exim filters).
2708 6. When Exim created an email address using the login of the caller as the
2709 local part (e.g. when creating a From: or Sender: header line), it was not
2710 quoting the local part when it contained special characters such as @.
2712 7. Installed new OpenBSD configuration files.
2714 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2715 try to make them clearer.
2717 9. Callout options, other than the timeout value, were being ignored when
2718 verifying sender addresses in header lines. For example, when using
2720 verify = header_sender/callout=no_cache
2722 the cache was (incorrectly) being used.
2724 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2725 only the Cygwin environment.
2727 11. When return_path_on_delivery was set as a log selector, if different remote
2728 addresses in the same message used different return paths and parallel
2729 remote delivery occurred, the wrong values would sometimes be logged.
2730 (Whenever a remote delivery process finished, the return path value from
2731 the most recently started remote delivery process was logged.)
2733 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2734 lines when AUTH and/or TLS are in use. This is the "received protocol"
2735 field. Exim used to use "asmtp" for authenticated SMTP, without any
2736 indication (in the protocol name) for TLS use. Now it follows the RFC and
2737 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2738 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2739 names appear in log lines as well as in Received: header lines.
2741 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2742 extensions, and comparison tests, and to fix some bugs.
2744 14. Changes to the "personal" filter test:
2746 (1) The test was buggy in that it was just doing the equivalent of
2747 "contains" tests on header lines. For example, if a user's address was
2748 anne@some.where, the "personal" test would incorrectly be true for
2750 To: susanne@some.where
2752 This test is now done by extracting each address from the header in turn,
2753 and checking the entire address. Other tests that are part of "personal"
2754 are now done using regular expressions (for example, to check local parts
2755 of addresses in From: header lines).
2757 (2) The list of non-personal local parts in From: addresses has been
2758 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2759 taken from the Sieve specification recommendations.
2761 (3) If the message contains any header line starting with "List-" it is
2762 treated as non-personal.
2764 (4) The test for "circular" in the Subject: header line has been removed
2765 because it now seems ill-conceived.
2767 15. Minor typos in src/EDITME comments corrected.
2769 16. Installed latest exipick from John Jetmore.
2771 17. If headers_add on a router specified a text string that was too long for
2772 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2773 of string_sprintf() is now avoided.
2775 18. $message_body_size was not set (it was always zero) when running the DATA
2776 ACL and the local_scan() function.
2778 19. For the "mail" command in an Exim filter, no default was being set for
2779 the once_repeat time, causing a random time value to be used if "once" was
2780 specified. (If the value happened to be <= 0, no repeat happened.) The
2781 default is now 0s, meaning "never repeat". The "vacation" command was OK
2782 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2783 (I found it when inspecting the code).
2785 20. There is now an overall timeout for performing a callout verification. It
2786 defaults to 4 times the callout timeout, which applies to individual SMTP
2787 commands during the callout. The overall timeout applies when there is more
2788 than one host that can be tried. The timeout is checked before trying the
2789 next host. This prevents very long delays if there are a large number of
2790 hosts and all are timing out (e.g. when the network connections are timing
2791 out). The value of the overall timeout can be changed by specifying an
2792 additional sub-option for "callout", called "maxwait". For example:
2794 verify = sender/callout=5s,maxwait=20s
2796 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2797 to the end before writing, but this should make it even safer).
2799 22. Exim was forgetting that it had advertised PIPELINING for the second and
2800 subsequent messages on an SMTP connection. It was also not resetting its
2801 memory on STARTTLS and an internal HELO.
2803 23. When Exim logs an SMTP synchronization error within a session, it now
2804 records whether PIPELINING has been advertised or not.
2806 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2807 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2808 rather than long int.
2810 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2812 26. Added the never_mail option to autoreply.
2818 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2819 crash if the getsockname() call failed; this can happen if a connection is
2820 closed very soon after it is established. The problem was simply in the
2821 order in which certain operations were done, causing Exim to try to write
2822 to the SMTP stream before it had set up the file descriptor. The bug has
2823 been fixed by making things happen in the correct order.
2829 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2830 before the connection was closed, thus losing the rejection response.
2832 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2833 some early Solaris releases, but causes trouble in current releases where
2834 socklen_t is defined.
2836 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2839 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2842 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2843 that releasing the top of it at the end releases what was used for sub-
2844 expansions (unless the block got too big). However, discard this block if
2845 the first thing is a variable or header, so that we can use its block when
2846 it is dynamic (useful for very large $message_headers, for example).
2848 6. Lookups now cache *every* query, not just the most recent. A new, separate
2849 store pool is used for this. It can be recovered when all lookup caches are
2850 flushed. Lookups now release memory at the end of their result strings.
2851 This has involved some general refactoring of the lookup sources.
2853 7. Some code has been added to the store_xxx() functions to reduce the amount
2854 of flapping under certain conditions.
2856 8. log_incoming_interface used to affect only the <= reception log lines. Now
2857 it causes the local interface and port to be added to several more SMTP log
2858 lines, for example "SMTP connection from", and rejection lines.
2860 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2862 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2864 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2865 could be overwritten at the end of the current message (or the start of a
2866 new message if it was set in a HELO ACL). The value is now preserved for
2867 the duration of the SMTP connection.
2869 12. If a transport had a headers_rewrite setting, and a matching header line
2870 contained an unqualified address, that address was qualified, even if it
2871 did not match any rewriting rules. The underlying bug was that the values
2872 of the flags that permit the existence of unqualified sender and recipient
2873 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2874 non-local messages, and by -bnq for local messages) were not being
2875 preserved with the message after it was received.
2877 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2878 "next input=" as part of the text comprising the host identity instead of
2879 the correct text. The code was using the same buffer for two different
2880 strings. However, depending on which order the printing function evaluated
2881 its arguments, the bug did not always show up. Under Linux, for example, my
2882 test suite worked just fine.
2884 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2885 doesn't work with some older versions of Perl. It has been changed to "my",
2886 which in any case is probably the better facility to use.
2888 15. A really picky compiler found some instances of statements for creating
2889 error messages that either had too many or two few arguments for the format
2892 16. The size of the buffer for calls to the DNS resolver has been increased
2893 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2894 for addresses that have a lot of PTR records. This alleviates a problem; it
2895 does not fully solve it.
2897 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2898 buffer now truncates the list and logs the incident, which is the same
2899 action as happens when Exim is looking up a host name and its aliases.
2900 Previously in this situation something unpredictable would happen;
2901 sometimes it was "internal error: store_reset failed".
2903 18. If a server dropped the connection unexpectedly when an Exim client was
2904 using GnuTLS and trying to read a response, the client delivery process
2905 crashed while trying to generate an error log message.
2907 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2908 string, for example:
2910 warn message = H1: something\nH2: something
2912 the text was added as a single header line from Exim's point of view
2913 though it ended up OK in the delivered message. However, searching for the
2914 second and subsequent header lines using $h_h2: did not work. This has been
2915 fixed. Similarly, if a system filter added multiple headers in this way,
2916 the routers could not see them.
2918 20. Expanded the error message when iplsearch is called with an invalid key to
2919 suggest using net-iplsearch in a host list.
2921 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2922 ACLs are no longer actually imposed (and a message to that effect is
2925 22. If a "gecos" field in a passwd entry contained escaped characters, in
2926 particular, if it contained a \" sequence, Exim got it wrong when building
2927 a From: or a Sender: header from that name. A second bug also caused
2928 incorrect handling when an unquoted " was present following a character
2929 that needed quoting.
2931 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2932 was not being matched caselessly.
2934 24. Arranged for all hyphens in the exim.8 source to be escaped with
2937 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2938 properly. Recipient callout cache records were still being keyed to include
2939 the sender, even when use_sender was set false. This led to far more
2940 callouts that were necessary. The sender is no longer included in the key
2941 when use_sender is false.
2943 26. Added "control = submission" modifier to ACLs.
2945 27. Added the ${base62d: operator to decode base 62 numbers.
2947 28. dnsdb lookups can now access SRV records.
2949 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2950 the configuration file.
2952 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2953 (-v) mode. This makes the output for a verbose queue run more intelligible.
2955 31. Added a use_postmaster feature to recipient callouts.
2957 32. Added the $body_zerocount variable, containing the number of binary zero
2958 bytes in the message body.
2960 33. The time of last modification of the "new" subdirectory is now used as the
2961 "mailbox time last read" when there is a quota error for a maildir
2964 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2966 35. Added +ignore_unknown as a special item in host lists.
2968 36. Code for decoding IPv6 addresses in host lists is now included, even if
2969 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2970 address was recognized as an IP address, but was then not correctly decoded
2971 into binary, causing unexpected and incorrect effects when compared with
2978 1. Very minor rewording of debugging text in manualroute to say "list of
2979 hosts" instead of "hostlist".
2981 2. If verify=header_syntax was set, and a header line with an unqualified
2982 address (no domain) and a large number of spaces between the end of the
2983 name and the colon was received, the reception process suffered a buffer
2984 overflow, and (when I tested it) crashed. This was caused by some obsolete
2985 code that should have been removed. The fix is to remove it!
2987 3. When running in the test harness, delay a bit after writing a bounce
2988 message to get a bit more predictability in the log output.
2990 4. Added a call to search_tidyup() just before forking a reception process. In
2991 theory, someone could use a lookup in the expansion of smtp_accept_max_
2992 per_host which, without the tidyup, could leave open a database connection.
2994 5. Added the variables $recipient_data and $sender_data which get set from a
2995 lookup success in an ACL "recipients" or "senders" condition, or a router
2996 "senders" option, similar to $domain_data and $local_part_data.
2998 6. Moved the writing of debug_print from before to after the "senders" test
3001 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3002 problems for message scanning, either using a data ACL, or using
3003 local_scan() because the Received: header was not generated till after they
3004 were called (in order to set the time as the time of reception completion).
3005 I have revised the way this works. The header is now generated after the
3006 body is received, but before the ACL or local_scan() are called. After they
3007 are run, the timestamp in the header is updated.
3013 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3014 before starting a queue runner without re-exec. This happened only when
3015 deliver_drop_privilege was set or when the Exim user was set to root. The
3016 effect of the bug was that timeouts during subsequent deliveries caused
3017 crashes instead of being properly handled. The handler is now left at its
3018 default (and expected) setting.
3020 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3021 message, again when deliver_drop_privilege is set or Exim is run as root.
3022 The bug described in (1) was not present in this case, but the tidying up
3023 of the other signals was missing. I have made the two cases consistent.
3025 3. The ignore_target_hosts setting on a manualroute router was being ignored
3026 for hosts that were looked up using the /MX notation.
3028 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3031 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3032 operated on the sender address. After changing the $sender_address to <>
3033 for the sender address verify, Exim was re-instated it as the original
3034 (before rewriting) address, but remembering that it had rewritten it, so it
3035 wasn't rewriting it again. This bug also had the effect of breaking the
3036 sender address verification caching when the sender address was rewritten.
3038 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3039 This has been changed so that if the ip literal address matches
3040 ignore_target_hosts, the router declines.
3042 7. Added expansion conditions match_domain, match_address, and match_local_
3043 part (NOT match_host).
3045 8. The placeholder for the Received: header didn't have a length field set.
3047 9. Added code to Exim itself and to exim_lock to test for a specific race
3048 condition that could lead to file corruption when using MBX delivery. The
3049 issue is with the lockfile that is created in /tmp. If this file is removed
3050 after a process has opened it but before that process has acquired a lock,
3051 there is the potential for a second process to recreate the file and also
3052 acquire a lock. This could lead to two Exim processes writing to the file
3053 at the same time. The added code performs the same test as UW imapd; it
3054 checks after acquiring the lock that its file descriptor still refers to
3055 the same named file.
3057 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3058 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3061 11. Added the smtp_active_hostname option. If used, this will typically be made
3062 to depend on the incoming interface address. Because $interface_address is
3063 not set up until the daemon has forked a reception process, error responses
3064 that can happen earlier (such as "too many connections") no longer contain
3067 12. If an expansion in a condition on a "warn" statement fails because a lookup
3068 defers, the "warn" statement is abandoned, and the next ACL statement is
3069 processed. Previously this caused the whole ACL to be aborted.
3071 13. Added the iplsearch lookup type.
3073 14. Added ident_timeout as a log selector.
3075 15. Added tls_certificate_verified as a log selector.
3077 16. Added a global option tls_require_ciphers (compare the smtp transport
3078 option of the same name). This controls incoming TLS connections.
3080 17. I finally figured out how to make tls_require_ciphers do a similar thing
3081 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3082 before starting the TLS session.
3084 18. Tabs are now shown as \t in -bP output.
3086 19. If the log selector return_path_on_delivery was set, Exim crashed when
3087 bouncing a message because it had too many Received: header lines.
3089 20. If two routers both had headers_remove settings, and the first one included
3090 a superfluous trailing colon, the final name in the first list and the
3091 first name in the second list were incorrectly joined into one item (with a
3092 colon in the middle).
3098 1. Added -C and -D options to the exinext utility, mainly to make it easier
3099 to include in the automated testing, but these could be helpful when
3100 multiple configurations are in use.
3102 2. The exinext utility was not formatting the output nicely when there was
3103 an alternate port involved in the retry record key, nor when there was a
3104 message id as well (for retries that were specific to a specific message
3105 and a specific host). It was also confused by IPv6 addresses, because of
3106 the additional colons they contain. I have fixed the IPv4 problem, and
3107 patched it up to do a reasonable job for IPv6.
3109 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3110 delivery, the log line now contains "pipelined" if PIPELINING was used.
3112 4. An SMTP transport process used to panic and die if the bind() call to set
3113 an explicit outgoing interface failed. This has been changed; it is now
3114 treated in the same way as a connect() failure.
3116 5. A reference to $sender_host_name in the part of a conditional expansion
3117 that was being skipped was still causing a DNS lookup. This no longer
3120 6. The def: expansion condition was not recognizing references to header lines
3121 that used bh_ and bheader_.
3123 7. Added the _cache feature to named lists.
3125 8. The code for checking quota_filecount in the appendfile transport was
3126 allowing one more file than it should have been.
3128 9. For compatibility with Sendmail, the command line option
3136 and sets the incoming protocol and host name (for trusted callers). The
3137 host name and its colon can be omitted when only the protocol is to be set.
3138 Note the Exim already has two private options, -pd and -ps, that refer to
3139 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3140 "s", but I don't think that's a major issue.
3142 10. A number of refactoring changes to the code, none of which should affect
3145 (a) The number of logging options was getting close to filling up the
3146 32-bit word that was used as a bit map. I have split them into two classes:
3147 those that are passed in the argument to log_write(), and those that are
3148 only ever tested independently outside of that function. These are now in
3149 separate 32-bit words, so there is plenty of room for expansion again.
3150 There is no change in the user interface or the logging behaviour.
3152 (b) When building, for example, log lines, the code previously used a
3153 macro that called string_cat() twice, in order to add two strings. This is
3154 not really sufficiently general. Furthermore, there was one instance where
3155 it was actually wrong because one of the argument was used twice, and in
3156 one call a function was used. (As it happened, calling the function twice
3157 did not affect the overall behaviour.) The macro has been replaced by a
3158 function that can join an arbitrary number of extra strings onto a growing
3161 (c) The code for expansion conditions now uses a table and a binary chop
3162 instead of a serial search (which was left over from when there were very
3163 few conditions). Also, it now recognizes conditions like "pam" even when
3164 the relevant support is not compiled in: a suitably worded error message is
3165 given if an attempt is made to use such a condition.
3167 11. Added ${time_interval:xxxxx}.
3169 12. A bug was causing one of the ddress fields not to be passed back correctly
3170 from remote delivery subprocesses. The field in question was not being
3171 subsequently used, so this caused to problems in practice.
3173 13. Added new log selectors queue_time and deliver_time.
3175 14. Might have fixed a bug in maildirsizefile handling that threw up
3176 "unexpected character" debug warnings, and recalculated the data
3177 unnecessarily. In any case, I expanded the warning message to give more
3180 15. Added the message "Restricted characters in address" to the statements in
3181 the default ACL that block characters like @ and % in local parts.
3183 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3184 Three changes have been made:
3186 (a) There was a serious bug; a negative response to MAIL caused the whole
3187 recipient domain to be cached as invalid, thereby blocking all messages
3188 to all local parts at the same domain, from all senders. This bug has
3189 been fixed. The domain is no longer cached after a negative response to
3190 MAIL if the sender used is not empty.
3192 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3195 (c) A new callout option, "use_sender" has been added for people who want
3196 the modified behaviour.
3202 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3205 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3206 indeed breaks things for older releases.
3208 3. Added additional logging to the case where there is a problem reading data
3209 from a filter that is running in a subprocess using a pipe, in order to
3210 try to track down a specific problem.
3212 4. Testing facility fudge: when running in the test harness and attempting
3213 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3214 getting "No route to host". Convert this to a timeout.
3216 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3219 6. Some OS don't have socklen_t but use size_t instead. This affects the
3220 fifth argument of getsockopt() amongst other things. This is now
3221 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3222 can be set for individual OS. I have set it for SunOS5, OSF1, and
3223 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3224 some earlier ones do not.
3226 7. Change 4.30/15 was not doing the test caselessly.
3228 8. The standard form for an IPv6 address literal was being rejected by address
3229 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3230 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3231 this, as well as the form without the "IPv6" on the front (but only when
3232 address literals are enabled, of course).
3234 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3236 10. Exim crashed if a message with an empty sender address specified by -f
3237 encountered a router with an errors_to setting. This could be provoked only
3238 by a command such as
3242 where an empty string was supplied; "<>" did not hit this bug.
3244 11. Installed PCRE release 4.5.
3246 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3247 remained set. It is now erased.
3249 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3250 times from message ids (which are base 36 rather than the normal 62).
3252 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3253 were being counted as actual protocol errors, and logged if the log
3254 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3255 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3256 and DATA following a set of rejected RCPTs do not count as protocol errors.
3257 In other words, Exim assumes they were pipelined, though this may not
3258 actually be the case. Of course, in all cases the client gets an
3259 appropriate error code.
3261 15. If a lookup fails in an ACL condition, a message about the failure may
3262 be available; it is used if testing the ACL cannot continue, because most
3263 such messages specify what the cause of the deferral is. However, some
3264 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3265 that caused an old message to be retained and used if a later statement
3266 caused a defer, replacing the real cause of the deferral.
3268 16. If an IP address had so many PTR records that the DNS lookup buffer
3269 was not large enough to hold them, Exim could crash while trying to process
3270 the truncated data. It now detects and logs this case.
3272 17. Further to 4.21/58, another change has been made: if (and only if) the
3273 first line of a message (the first header line) ends with CRLF, a bare LF
3274 in a subsequent header line has a space inserted after it, so as not to
3275 terminate the header.
3277 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3278 unnecessarily, used atoi() instead of strtol(), and didn't check the
3279 termination when getting file sizes from file names by regex.
3281 19. Completely re-implemented the support for maildirsize files, in the light
3282 of a number of problems with the previous contributed implementation
3283 (4.30/29). In particular:
3285 . If the quota is zero, the maildirsize file is maintained, but no quota is
3288 . If the maildir directory does not exist, it is created before any attempt
3289 to write a maildirsize file.
3291 . The quota value in the file is just a cache; if the quota is changed in
3292 the transport, the new value overrides.
3294 . A regular expression is available for excluding directories from the
3297 20. The autoreply transport checks the characters in options that define the
3298 message's headers; it allows continued headers, but it was checking with
3299 isspace() after an embedded newline instead of explicitly looking for a
3302 21. If all the "regular" hosts to which an address was routed had passed their
3303 expiry times, and had not reached their retry times, the address was
3304 bounced, even if fallback hosts were defined. Now Exim should go on to try
3307 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3308 equivalent code in the SMTP transport. Some hosts send humungous responses
3309 to HELO/EHLO, more than 1024 it seems.
3311 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3312 gives compiler warnings in some environments. I've now done it "properly",
3315 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3316 (because of problems with the built-in one) was declared to return uschar *
3317 instead of char *, causing compiler failure.
3319 25. Fixed a file descriptor leak when processing alias/forward files.
3321 26. Fixed a minor format string issue in dbfn.c.
3323 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3325 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3326 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3329 29. When a sender address is verified, it is cached, to save repeating the test
3330 when there is more than one recipient in a message. However, when the
3331 verification involves a callout, it is possible for different callout
3332 options to be set for different recipients. It is too complicated to keep
3333 track of this in the cache, so now Exim always runs a verification when a
3334 callout is required, relying on the callout cache for the optimization.
3335 The overhead is duplication of the address routing, but this should not be
3338 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3339 to be verified with callout=postmaster, and the main callout worked but the
3340 postmaster check failed, the verification correctly failed. However, if a
3341 subsequent RCPT command asked for sender verification *without* the
3342 postmaster check, incorrect caching caused this verification also to fail,
3345 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3346 it was not caching the DNS options (qualify_single, search_parents) that
3347 were used when the lookup failed. A subsequent lookup with different
3348 options therefore always gave the same answer, though there were cases
3349 where it should not have. (Example: a "domains = !$mx_any" option on a
3350 dnslookup router: the "domains" option is always processed without any
3351 widening, but the router might have qualify_single set.) Now Exim uses the
3352 cached value only when the same options are set.
3354 32. Added John Jetmore's "exipick" utility to the distribution.
3356 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3357 than a timeout (e.g. a certificate is required, and is not provided), an
3358 Exim server now closes the connection immediately. Previously it waited for
3359 the client to close - but if the client is SSL, it seems that they each
3360 wait for each other, leading to a delay before one of them times out.
3362 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3363 maintained 0.8.x compatibility because I don't think many are using it, and
3364 it is clearly obsolete.
3366 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3369 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3370 client certificate was expired. A simple patch fixes this, though I don't
3371 understand the full logic of why the verify callback is called multiple
3374 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3375 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3376 which causes problems with some clients (such as the Certicom SSL Plus
3377 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3378 disables the coutermeasure allowing Eudora to connect."
3380 38. Exim was not checking that a write() to a log file succeeded. This could
3381 lead to Bad Things if a log got too big, in particular if it hit a file
3382 size limit. Exim now panics and dies if it cannot write to a log file, just
3383 as it does if it cannot open a log file.
3385 39. Modified OS/Makefile-Linux so that it now contains
3387 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3389 The two -D definitions ensure that Exim is compiled with large file
3390 support, which makes it possible to handle log files that are bigger than
3393 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3394 instance) a domain was checked against a named list that involved a lookup,
3395 causing $domain_data to be set, then another domain was checked against the
3396 same list, then the first domain was re-checked, the value of $domain_data
3397 after the final check could be wrong. In particular, if the second check
3398 failed, it could be set empty. This bug probably also applied to
3401 41. The strip_trailing_dot option was not being applied to the address given
3402 with the -f command-line option.
3404 42. The code for reading a message's header from the spool was incrementing
3405 $received_count, but never initializing it. This meant that the value was
3406 incorrect (doubled) while delivering a message in the same process in which
3407 it was received. In the most common configuration of Exim, this never
3408 happens - a fresh exec is done - but it can happen when
3409 deliver_drop_privilege is set.
3411 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3412 it now includes up to 150 characters of the unexpected data in the log
3415 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3416 and building data strings. The size of both of these buffers was 10 000
3417 bytes - far larger than anybody would *ever* want, thought I. Needless to
3418 say, somebody hit the limit. I have increased the maximum line length to
3419 20 000 and the maximum data length of concatenated lines to 100 000. I have
3420 also fixed two bugs, because there was no checking on these buffers. Tsk,
3421 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3422 buffer is too small.
3424 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3425 lsearch lookups. Now it does.
3427 46. When parsing a route_list item in a manualroute router, a fixed-length
3428 buffer was used for the list of hosts. I made this 1024 bytes long,
3429 thinking that nobody would ever have a list of hosts that long. Wrong.
3430 Somebody had a whole pile of complicated expansion conditions, and the
3431 string was silently truncated, leading to an expansion error. It turns out
3432 that it is easier to change to an unlimited length (owing to other changes
3433 that have happened since this code was originally written) than to build
3434 structure for giving a limitation error. The length of the item that
3435 expands into the list of hosts is now unlimited.
3437 47. The lsearch lookup could not handle data where the length of text line was
3438 more than 4095 characters. Such lines were truncated, leading to shortened
3439 data being returned. It should now handle lines of any length.
3441 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3442 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3445 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3446 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3447 no longer happens in comment lines. A list of replacements is now placed
3448 at the head of all of the source files, except those whose only change is
3449 to replace PERL_COMMAND in the very first #! line.
3451 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3452 messages on the queue, with a bottom-up merge sort, using code contributed
3453 by Michael Haardt. This should make operations like -bp somewhat faster on
3454 large queues. It won't affect queue runners, except when queue_run_in_order
3457 51. Installed eximstats 1.31 in the distribution.
3459 52. Added support for SRV lookups to the dnslookup router.
3461 53. If an ACL referred to $message_body or $message_body_end, the value was not
3462 reset for any messages that followed in the same SMTP session.
3464 54. The store-handling optimization for building very long strings was not
3465 differentiating between the different store pools. I don't think this
3466 actually made any difference in practice, but I've tidied it.
3468 55. While running the routers to verify a sender address, $sender_address
3469 was still set to the sender address. This is wrong, because when routing to
3470 send a bounce to the sender, it would be empty. Therefore, I have changed
3471 it so that, while verifying a sender address, $sender_address is set to <>.
3472 (There is no change to what happens when verifying a recipient address.)
3474 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3475 target A or AAAA records (if not already returned) without resetting the
3476 qualify_single or search_parents options of the DNS resolver. These are
3477 inappropriate in this case because the targets of MX and SRV records must
3478 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3479 target that, when qualified, matched something in the local domain. These
3480 two options are now turned off when doing these lookups.
3482 57. It seems that at least some releases of Reiserfs (which does not have the
3483 concept of a fixed number of inodes) returns zero and not -1 for the
3484 number of available inodes. This interacted badly with check_spool_inodes,
3485 which assumed that -1 was the "no such thing" setting. What I have done is
3486 to check that the total number of inodes is greater than zero before doing
3487 the test of how many are available.
3489 58. When a "warn" ACL statement has a log_message modifier, the message is
3490 remembered, and not repeated. This is to avoid a lot of repetition when a
3491 message has many recipients that cause the same warning to be written.
3492 Howewer, Exim was preserving the list of already written lines for an
3493 entire SMTP session, which doesn't seem right. The memory is now reset if a
3494 new message is started.
3496 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3497 debugging output unless log_rewrite was also set.
3499 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3500 of (int)(handle) when we know that handle contains (void *)(-1).
3502 61. The Exim daemon panic-logs an error return when it closes the incoming
3503 connection. However "connection reset by peer" seems to be common, and
3504 isn't really an error worthy of noting specially, so that particular error
3507 62. When Exim is trying to find all the local interfaces, it used to panic and
3508 die if the ioctl to get the interface flags failed. However, it seems that
3509 on at least one OS (Solaris 9) it is possible to have an interface that is
3510 included in the list of interfaces, but for which you get a failure error
3511 for this call. This happens when the interface is not "plumbed" into a
3512 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3513 failure of the "get flags" call assumes that the interface is down.
3515 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3516 makes life easier for people who are doing arithmetic on fields extracted
3517 from dates, where you often get leading zeros that should not be
3518 interpreted as octal.
3520 64. Added qualify_domain to the redirect router, to override the global
3523 65. If a pathologically long header line contained very many addresses (the
3524 report of this problem mentioned 10 000) and each of them was rewritten,
3525 Exim could use up a very large amount of memory. (It kept on making new
3526 copies of the header line as it rewrote, and never released the old ones.)
3527 At the expense of a bit more processing, the header rewriting function has
3528 been changed so that it no longer eats memory in this way.
3530 66. The generation of the Received: header has been moved from the time that a
3531 message starts to be received, to the time that it finishes. The timestamp
3532 in the Received: header should now be very close to that of the <= log
3533 line. There are two side-effects of this change:
3535 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3536 logged header lines no longer include the local Received: line, because
3537 it has not yet been created. The same applies to a copy of the message
3538 that is returned to a non-SMTP sender when a message is rejected.
3540 (b) When a filter file is tested using -bf, no additional Received: header
3541 is added to the test message. After some thought, I decided that this
3544 This change does not affect the value of $received_for. It is still set
3545 after address rewriting, but before local_scan() is called.
3547 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3549 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3550 gave an unhelpful panic error message, and a defer error. I have managed to
3551 change this behaviour so that it now rejects any supplied certificate,
3552 which seems right, as the list of acceptable certificates is empty.
3554 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3555 gave an unhelpful defer error. I have not managed to make this reject any
3556 supplied certificates, but the error message it gives is "no certificate
3557 supplied", which is not helpful.
3559 70. exigrep's output now also includes lines that are not associated with any
3560 message, but which match the given pattern. Implemented by a patch from
3561 Martin Sluka, which also tidied up the Perl a bit.
3563 71. Recipient callout verification, like sender verification, was using <> in
3564 the MAIL FROM command. This isn't really the right thing, since the actual
3565 sender may affect whether the remote host accepts the recipient or not. I
3566 have changed it to use the actual sender in the callout; this means that
3567 the cache record is now keyed on a recipient/sender pair, not just the
3568 recipient address. There doesn't seem to be a real danger of callout loops,
3569 since a callout by the remote host to check the sender would use <>.
3570 [SEE ABOVE: changed after hitting problems.]
3572 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3573 temporary errors. However, in the case of such a code being given after
3574 the end of a data transmission (i.e. after ".") Exim was failing to write
3575 a retry record for the message. (Yes, there was some broken host that was
3576 actually sending 8xx at this point.)
3578 73. An unknown lookup type in a host list could cause Exim to panic-die when
3579 the list was checked. (An example that provoked this was putting <; in the
3580 middle of a list instead of at the start.) If this happened during a DATA
3581 ACL check, a -D file could be left lying around. This kind of configuration
3582 error no longer causes Exim to die; instead it causes a defer errror. The
3583 incident is still logged to the main and panic logs.
3585 74. Buglet left over from Exim 3 conversion. The message "too many messages
3586 in one connection" was written to the rejectlog but not the mainlog, except
3587 when address rewriting (yes!) was being logged.
3589 75. Added write_rejectlog option.
3591 76. When a system filter was run not as root (that is, when system_filter_user
3592 was set), the values of the $n variables were not being returned to the
3593 main process; thus, they were not subsequently available in the $sn
3596 77. Added +return_path_on_delivery log selector.
3598 78. A connection timeout was being treated differently from recipients deferred
3599 when testing hosts_max_try with a message that was older than the host's
3600 retry timeout. (The host should not be counted, thus allowing all hosts to
3601 be tried at least once before bouncing.) This may have been the cause of an
3602 occasionally reported bug whereby a message would remain on the queue
3603 longer than the retry timeout, but would be bounced if a delivery was
3604 forced. I say "may" because I never totally pinned down the problem;
3605 setting up timeout/retry tests is difficult. See also the next item.
3607 79. The ultimate address timeout was not being applied to errors that involved
3608 a combination of host plus message (for example, a timeout on a MAIL
3609 command). When an address resolved to a number of possible hosts, and they
3610 were not all tried for each delivery (e.g. because of hosts_max_try), a
3611 message could remain on the queue longer than the retry timeout.
3613 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3616 81. Fixed an obscure SMTP outgoing bug which required at least the following
3617 conditions: (a) there was another message waiting for the same server;
3618 (b) the server returned 5xx to all RCPT commands in the first message so
3619 that the message was not completed; (c) the server dropped the connection
3620 or gave a negative response to the RSET that Exim sends to abort the
3621 transaction. The observed case was a dropped connection after DATA that had
3622 been sent in pipelining mode. That is, the server had advertised PIPELINING
3623 but was not implementing it correctly. The effect of the bug was incorrect
3624 behaviour, such as trying another host, and this could lead to a crash.
3630 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3631 and daemon.c were passed as pointers to ints; they should have been
3632 pointers to socklen_t variables (which are typically unsigned ints).
3634 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3637 3. Fixed a really odd bug that affected only the testing scheme; patching a
3638 certain fixed string in the binary changed the value of another string that
3639 happened to be identical to the end of the original first string.
3641 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3642 name", it returns that address as the IP address. On some operating
3643 systems (e.g. Solaris), it also passes back the IP address string as the
3644 "host name". However, on others (e.g. Linux), it passes back an empty
3645 string. Exim wasn't checking for this, and was changing the host name to an
3646 empty string, assuming it had been canonicized.
3648 5. Although rare, it is permitted to have more than one PTR record for a given
3649 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3650 all the names associated with an address, because they do in Solaris.
3651 However, it seems that they do not in Linux for data that comes from the
3652 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3653 I found this out when I moved to a new Linux workstation and tried to run
3654 the Exim test suite.
3656 To get round this problem I have changed the code so that it now does its
3657 own call to the DNS to look up PTR records when searching for a host name.
3658 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3659 addresses that are only in /etc/hosts are still found.
3661 This behaviour is, however, controlled by an option called host_lookup_
3662 order, which defaults to "bydns:byaddr". If people want to use the other
3663 order, or indeed, just use one or the other means of lookup, they can
3664 specify it in this variable.
3666 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3667 some operating systems, this comes back from gethostbyaddr() as an empty
3668 string, and this is what Exim used to test for. However, it seems that in
3669 other systems, "." is yielded. Exim now tests for this case too.
3671 7. The values of check_spool_space and check_log_space are now held internally
3672 as a number of kilobytes instead of an absolute number of bytes. If a
3673 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3674 kilobyte. This means that much larger values can be stored.
3676 8. Exim monitor: an attempt to get the action menu when not actually pointing
3677 at a message produces an empty menu entitled "No message selected". This
3678 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3679 no entries in it ("Shell widget menu has zero width and/or height"). So I
3680 have added a single, blank menu entry in this case.
3682 9. Added ${quote_local_part.
3684 10. MIME decoding is now applied to the contents of Subject: header lines when
3687 11. Now that a reference to $sender_host_address automatically causes a reverse
3688 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3689 host lookup before query-style lookups in lists that might use this
3690 variable. This has therefore been abolished, and the "net-" prefix is no
3691 longer necessary for query-style lookups.
3693 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3694 have been a typo for LFLAGS, so it has been changed.
3696 13. The install script calls Exim with "-C /dev/null" in order to find the
3697 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3698 to be output. Howeve, since Exim outputs its version number before the
3699 error, it didn't break the script. It just looked ugly. I fixed this by
3700 always allowing "-C /dev/null" if the caller is root.
3702 14. Ignore overlarge ACL variable number when reading spool file - insurance
3703 against a later release with more variables having written the file.
3705 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3706 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3707 this, as well as the form without the "IPv6" on the front.
3709 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3710 OS/Makefile-Darwin file.
3712 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3713 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3715 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3716 variable. However, some OS use ints for inodes. I've added cast to long int
3717 to get rid of the compiler warning.
3719 19. I had forgotten to lock out "/../" in configuration file names when
3720 ALT_CONFIG_PREFIX was set.
3722 20. Routers used for verification do not need to specify transports. However,
3723 if such a router generated a host list, and callout was configured, Exim
3724 crashed, because it could not find a port number from the (non-existent)
3725 transport. It now assumes port 25 in this circumstance.
3727 21. Added the -t option to exigrep.
3729 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3730 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3731 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3732 them is set, LOOKUP_LSEARCH is forced.
3734 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3736 24. Added sender and host information to the "rejected by local_scan()" log
3737 line; previously there was no indication of these.
3739 25. Added .include_if_exists.
3741 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3742 Linux. It turns out that not all file systems support this. Apparently some
3743 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3744 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3745 not supported on the file descriptor, is now ignored when Exim is trying to
3746 sync a directory. This applies only to Linux.
3748 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3750 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3751 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3754 29. Added support for maildirsize files from supplied patch (modified a bit).
3756 30. The use of :fail: followed by an empty string could lead Exim to respond to
3757 sender verification failures with (e.g.):
3759 550 Verification failed for <xxx>
3760 550 Sender verify failed
3762 where the first response line was missing the '-' that indicates it is not
3763 the final line of the response.
3765 31. The loop for finding the name of the user that called Exim had a hardwired
3766 limit of 10; it now uses the value of finduser_retries, which is used for
3767 all other user lookups.
3769 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3772 33. Exim was neglecting to zero errno before one call of strtol() when
3773 expanding a string and expecting an integer value. On some systems this
3774 resulted in spurious "integer overflow" errors. Also, it was casting the
3775 result into an int without checking.
3777 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3778 did not work. The code looks as if it has *never* worked, though it appears
3779 to have been documented since at least releast 1.62. I have made it work.
3781 35. The "timeout_DNS" error in retry rules, also documented since at least
3782 1.62, also never worked. As it isn't clear exactly what this means, and
3783 clearly it isn't a major issue, I have abolished the feature by treating it
3784 as "timeout", and writing a warning to the main and panic logs.
3786 36. The display of retry rules for -brt wasn't always showing the error code
3789 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3790 timeout_connect_A, timeout_connect_MX.
3792 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3793 to the empty sender.
3795 39. The daemon was not analysing the content of -oX till after it had closed
3796 stderr and disconnected from the controlling terminal. This meant that any
3797 syntax errors were only noted on the panic log, and the return code from
3798 the command was 0. By re-arranging the code a little, I've made the
3799 decoding happen first, so such errors now appear on stderr, and the return
3800 code is 1. However, the actual setting up of the sockets still happens in
3801 the disconnected process, so errors there are still only recorded on the
3804 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3805 connections (as happens on some IP stacks) was logged at start up time as
3806 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3807 it from "IPv6 and IPv4", which means that two separate sockets are being
3810 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3811 says whether AF_INET or AF_INET6 was passed as an argument.
3813 42. Exiwhat output was messed up when time zones were included in log
3816 43. Exiwhat now gives more information about the daemon's listening ports,
3817 and whether -tls-on-connect was used.
3819 44. The "port" option of the smtp transport is now expanded.
3821 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3822 silently ignored. Now an error message is written to the main and panic
3825 46. There's a new ACL modifier called "logwrite" which writes to a log file
3826 as soon as it is encountered.
3828 47. Added $local_user_uid and $local_user_gid at routing time.
3830 48. Exim crashed when trying to verify a sender address that was being
3833 49. Exim was recognizing only a space character after ".include". It now also
3834 recognizes a tab character.
3836 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3837 extracting the relevant information from the specification. The man page no
3838 longer contains scrambled data for the -d option, and I've added a section
3839 at the front about calling Exim under different names.
3841 51. Added "extra_headers" argument to the "mail" command in filter files.
3843 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3846 53. Installed eximstats 1.29.
3848 54. Added transport_filter_timeout as a generic transport option.
3850 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3851 Cc: header lines. This was required by RFC 822, but it not required by RFC
3854 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3855 incoming messages that did not have them. Now it does so only if the
3856 message originates locally, that is, if there is no associated remote host
3857 address. When Resent- header lines are present, this applies to the Resent-
3858 lines rather than the non-Resent- lines.
3860 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3861 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3863 58. Messages for configuration errors now include the name of the main
3864 configuration file - useful now that there may be more than one file in a
3865 list (.included file names were always shown).
3867 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3868 for those rare installations that do not start the daemon as root or run it
3869 setuid root. I've cut out the call to initgroups() if the daemon is not
3872 60. The Exim user and group can now be bound into the binary as text strings
3873 that are looked up at the start of Exim's processing.
3875 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3877 62. Added $mailstore_basename variable.
3879 63. Installed patch to sieve.c from Michael Haardt.
3881 64. When Exim failed to open the panic log after failing to open the main log,
3882 the original message it was trying to log was written to stderr and debug
3883 output, but if they were not available (the usual case in production), it
3884 was lost. Now it is written to syslog before the two lines that record the
3885 failures to open the logs.
3887 65. Users' Exim filters run in subprocesses under the user's uid. It is
3888 possible for a "deliver" command or an alias in a "personal" command to
3889 provoke an address rewrite. If logging of address rewriting is configured,
3890 this fails because the process is not running as root or exim. There may be
3891 a better way of dealing with this, but for the moment (because 4.30 needs
3892 to be released), I have disabled address rewrite logging when running a
3893 filter in a non-root, non-exim process.
3899 1. The buildconfig auxiliary program wasn't quoting the value set for
3900 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3901 not defined. This bug was masked in 4.22 by the effect that was fixed in
3904 2. Some messages that were rejected after a message id was allocated were
3905 shown as "incomplete" by exigrep. It no longer does this for messages that
3906 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3908 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3909 have allow_domain_literals set, the ID did not get logged in the <= line.
3910 Domain literals are now always recognized in Message-ID: header lines.
3912 4. The first argument for a ${extract expansion item is the key name or field
3913 number. Leading and trailing spaces in this item were not being ignored,
3914 causing some misleading effects.
3916 5. When deliver_drop_privilege was set, single queue runner processes started
3917 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3918 same command in the process it spins off) were not dropping privilege.
3920 6. When the daemon running as "exim" started a queue runner, it always
3921 re-executed Exim in the spun-off process. This is a waste of effort when
3922 deliver_drop_privilege is set. The new process now just calls the
3923 queue-runner function directly.
3929 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3932 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3933 ignored. Though the use of -oP was forcing the writing of a pid file, it
3934 was always written to the default place.
3936 3. If the message "no IP address found for host xxxx" is generated during
3937 incoming verification, it is now followed by identification of the incoming
3938 connection (so you can more easily find what provoked it).
3940 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3942 5. Added some features to "harden" Exim a bit more against certain attacks:
3944 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3945 be put in Local/Makefile. This is like the never_users runtime option,
3946 but it cannot be overridden. The default setting is "root".
3948 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3949 prefix string with which any file named in a -C command line option
3952 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3953 is retained for -C and -D only if the caller of Exim is root. Without
3954 it, the exim user may also use -C and -D and retain privilege.
3956 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3957 command line option is disabled.
3959 6. Macro names set by the -D option must start with an upper case letter, just
3960 like macro names defined in the configuration file.
3962 7. Added "dereference=" facility to LDAP.
3964 8. Two instances of the typo "uknown" in the source files are fixed.
3966 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3967 the Configure-Makefile script screwed up while processing it.
3969 10. Incorporated PCRE 4.4.
3971 11. The SMTP synchronization check was not operating right at the start of an
3972 SMTP session. For example, it could not catch a HELO sent before the client
3973 waited for the greeting. There is now a check for outstanding input at the
3974 point when the greeting is written. Because of the duplex, asynchronous
3975 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3976 way, but not yet received, when the check is performed.
3978 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3979 on TCP/IP sockets, because this apparently causes some broken clients to
3982 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3983 unchanged) from the Cygwin maintainer.
3985 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3986 was supported instead of testing for mbx. Effectively a typo.
3988 15. The spa authenticator server code was not checking that the input it
3989 received was valid base64.
3991 16. The debug output line for the "set" modifier in ACLs was not showing the
3992 name of the variable that was being set.
3994 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3996 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3997 lookup on its own. Something else was needed to trigger the lookup. For
3998 example, a match in host_lookup or the need for a host name in a host list.
3999 Now, if $sender_host_name is referenced and the host name has not yet been
4000 looked up, a lookup is performed. If the lookup fails, the variable remains
4001 empty, and $host_lookup_failed is set to "1".
4003 19. Added "eqi" as a case-independent comparison operator.
4005 20. The saslauthd authentication condition could segfault if neither service
4006 nor realm was specified.
4008 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4009 error message that was logged was misleading, and incoming SMTP
4010 connections were dropped. The message is now more accurate, and temporary
4011 errors are given to SMTP connections.
4013 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4014 (except RSET) with a 421 error, until QUIT is received. However, it was
4015 failing to send a response to QUIT.
4017 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4018 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4019 ACL is now run after the helo_try_verify_hosts code.
4021 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4022 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4023 case-independent, so other case variants are also recognized). Apparently
4024 some systems use these upper case variants.
4026 25. If more than two messages were waiting for the same host, and a transport
4027 filter was specified for the transport, Exim sent two messages over the
4028 same TCP/IP connection, and then failed with "socket operation on non-
4029 socket" when it tried to send the third.
4031 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4033 27. The extern definition of crypt16() in expand.c was not being excluded when
4034 the OS had its own crypt16() function.
4036 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4037 as a preferred synonym for return_size_limit, both as an option and as an
4040 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4042 30. Changed the default configuration ACL to relax the local part checking rule
4043 for addresses that are not in any local domains. For these addresses,
4044 slashes and pipe symbols are allowed within local parts, but the sequence
4045 /../ is explicitly forbidden.
4047 31. SPA server authentication was not clearing the challenge buffer before
4050 32. log_message in a "warn" ACL statement was writing to the reject log as
4051 well as to the main log, which contradicts the documentation and doesn't
4052 seem right (because no rejection is happening). So I have stopped it.
4054 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4055 However, I am unable to do any testing of this.
4057 34. Fixed an infelicity in the appendfile transport. When checking directories
4058 for a mailbox, to see if any needed to be created, it was accidentally
4059 using path names with one or more superfluous leading slashes; tracing
4060 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4062 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4063 contents are added to the log line that is written for every discarded
4064 recipient. (Previously a log_message setting was ignored.)
4066 36. The ${quote: operator now quotes the string if it is empty.
4068 37. The install script runs exim in order to find its version number. If for
4069 some reason other than non-existence or emptiness, which it checks, it
4070 could not run './exim', it was installing it with an empty version number,
4071 i.e. as "exim-". This error state is now caught, and the installation is
4074 38. An argument was missing from the function that creates an error message
4075 when Exim fails to connect to the socket for saslauthd authentication.
4076 This could cause Exim to crash, or give a corrupted message.
4078 39. Added isip, isip4, and isip6 to ${if conditions.
4080 40. The ACL variables $acl_xx are now saved with the message, and can be
4081 accessed later in routers, transports, and filters.
4083 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4084 strings in the file are not string-expanded.
4086 42. If a MAIL command specified a SIZE value that was too large to fit into an
4087 int variable, the check against message_size_limit failed. Such values are
4088 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4089 day this will have to be increased, but I don't think I want to be around
4090 when emails are that large.
4097 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4098 iconv() is not standard in FreeBSD.
4100 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4101 IPv6 enabled. The observed symptom was a segmentation fault on return from
4102 the function os_common_find_running_interfaces() in src/os.c.
4104 3. In the check_special_case() function in daemon.c I had used "errno" as an
4105 argument name, which causes warnings on some systems. This was basically a
4106 typo, since it was named "eno" in the comments!
4108 4. The code that waits for the clock to tick (at a resolution of some fraction
4109 of a second) so as to ensure message-id uniqueness was always waiting for
4110 at least one whole tick, when it could have waited for less. [This is
4111 almost certainly not relevant at current processor speeds, where it is
4112 unlikely to ever wait at all. But we try to future-proof.]
4114 5. The function that sleeps for a time interval that includes fractions of a
4115 second contained a race. It did not block SIGALRM between setting the
4116 timer, and suspending (a couple of lines later). If the interval was short
4117 and the sigsuspend() was delayed until after it had expired, the suspension
4118 never ended. On busy systems this could lead to processes getting stuck for
4121 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4122 process, before it forks any delivery processes. The open lookup caching
4123 mechanism meant that the open file or database connection was passed into
4124 the delivery process. The problem was that delivery processes always tidy
4125 up cached lookup data. This could cause a problem for the next delivery
4126 process started by the queue runner, because the external queue runner
4127 process does not know about the closure. So the next delivery process
4128 still has data in the lookup cache. In the case of a file lookup, there was
4129 no problem because closing a file descriptor in a subprocess doesn't affect
4130 the parent. However, if the lookup was caching a connection to a database,
4131 the connection was closed, and the second delivery process was likely to
4132 see errors such as "PGSQL: query failed: server closed the connection
4133 unexpectedly". The problem has been fixed by closing all cached lookups
4134 in a queue runner before running a delivery process.
4136 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4137 seem to have the "const" qualifier which it has on other OS. I've
4140 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4141 *of the same type* (client or server) with the same public name that an
4142 error should be diagnosed.
4144 9. When Exim looked up a host name for an IP address, but failed to find the
4145 original IP address when looking up the host name (a safety check), it
4146 output the message "<ip address> does not match any IP for NULL", which was
4147 confusing, to say the least. The bug was that the host name should have
4148 appeared instead of "NULL".
4150 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4151 user that is built into the binary, and the -C or -D options is used, root
4152 privilege is dropped before the configuration file is read. In addition,
4153 logging is switched to stderr instead of the normal log files. If the
4154 configuration then re-defines the Exim user, the unprivileged environment
4155 is probably not what is expected, so Exim logs a panic warning message (but
4158 However, if deliver_drop_privilege is set, the unprivileged state may well
4159 be exactly what is intended, so the warning has been cut out in that case,
4160 and Exim is allowed to try to write to its normal log files.
4166 1. smtp_return_error_details was not giving details for temporary sender
4167 or receiver verification errors.
4169 2. Diagnose a configuration error if two authenticators have the same public
4172 3. Exim used not to create the message log file for a message until the first
4173 delivery attempt. This could be confusing when incoming messages were held
4174 for policy or load reasons. The message log file is now created at the time
4175 the message is received, and an initial "Received" line is written to it.
4177 4. The automatically generated man page for command line options had a minor
4178 bug that caused no ill effects; however, a more serious problem was that
4179 the procedure for building the man page automatically didn't always
4180 operate. Consequently, release 4.20 contains an out-of-date version. This
4181 shouldn't happen again.
4183 5. When building Exim with embedded Perl support, the script that builds the
4184 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4185 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4187 6. The freeze_tell option was not being used for messages that were frozen on
4188 arrival, either by an ACL or by local_scan().
4190 7. Added the smtp_incomplete_transaction log selector.
4192 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4193 was accepting AUTH without a new EHLO.
4195 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4196 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4199 10. Logging of TCP/IP connections (when configured) now happens in the main
4200 daemon process instead of the child process, so that the TCP/IP connection
4201 count is more accurate (but it can never be perfect).
4203 11. The use of "drop" in a nested ACL was not being handled correctly in the
4204 outer ACL. Now, if condition failure induced by the nested "drop" causes
4205 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4206 or "require"), the connection is dropped.
4208 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4209 that yield "discard" can now be used with an "accept" or a "discard" verb,
4210 but an error is generated for any others (because I can't see a useful way
4211 to define what should happen).
4213 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4214 are now processed in the same way as lines in the Exim configuration file.
4215 In particular, continuation lines are supported.
4217 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4219 15. Added -ti meaning -t -i.
4221 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4222 domains, and warn by logging if others are found.
4224 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4225 returned from GIFCONF when Exim is trying to find the list of interfaces on
4226 a host. The code in os.c has been modified to copy each ifreq to an aligned
4227 structure in all cases.
4229 Also, in some cases, the returned ifreq's were being copied to a 'struct
4230 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4231 means the last couple of bytes of an IPv6 address could be chopped if the
4232 ifreq contained only a normal sockaddr (14 bytes storage).
4234 18. Named domain lists were not supported in the hosts_treat_as_local option.
4235 An entry such as +xxxx was not recognized, and was treated as a literal
4238 19. Ensure that header lines added by a DATA ACL are included in the reject log
4239 if the ACL subsequently rejects the message.
4241 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4242 MD5 (which is deprecated).
4244 21. When testing a filter file using -bf, Exim was writing a message when it
4245 took the sender from a "From " line in the message, but it was not doing so
4246 when it took $return_path from a Return-Path: header line. It now does.
4248 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4249 with a valid header line field name (a series of printing characters
4250 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4252 23. Changed "disc" in the source to "disk" to conform to the documentation and
4253 the book and for uniformity.
4255 24. Ignore Sendmail's -Ooption=value command line item.
4257 25. When execve() failed while trying to run a command in a pipe transport,
4258 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4259 could be confused with a return value of 69 from the command itself. This
4260 has been changed to 127, the value the shell returns if it is asked to run
4261 a non-existent command. The wording for the related log line suggests a
4262 non-existent command as the problem.
4264 26. If received_header_text expands to an empty string, do not add a Received:
4265 header line to the message. (Well, it adds a token one on the spool, but
4266 marks it "old" so that it doesn't get used or transmitted.)
4268 27. Installed eximstats 1.28 (addition of -nt option).
4270 28. There was no check for failure on the call to getsockname() in the daemon
4271 code. This can fail if there is a shortage of resources on the system, with
4272 ENOMEM, for example. A temporary error is now given on failure.
4274 29. Contrary to the C standard, it seems that in some environments, the
4275 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4276 program. Exim now does this explicitly; it affects the formatting of
4277 timestamps using strftime().
4279 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4280 complaints. I've now initialized all the variables, to avoid this.
4282 32. Header lines added by a system filter were not being "seen" during
4283 transport-time rewrites.
4285 33. The info_callback() function passed to OpenSSL is set up with type void
4286 (*)(SSL *, int, int), as described somewhere. However, when calling the
4287 function (actually a macro) that sets it up, the type void(*)() is
4288 expected. I've put in a cast to prevent warnings from picky compilers.
4290 34. If a DNS black list lookup found a CNAME record, but there were no A
4291 records associated with the domain it pointed at, Exim crashed.
4293 35. If a DNS black list lookup returned more than one A record, Exim ignored
4294 all but the first. It now scans all returned addresses if a particular IP
4295 value is being sought. In this situation, the contents of the
4296 $dnslist_value variable are a list of all the addresses, separated by a
4299 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4300 to accept a host name and all its aliases if the forward lookup for any of
4301 them yielded the IP address of the incoming connection. Now it accepts only
4302 those names whose forward lookup yields the correct IP address. Any other
4303 names are discarded. This closes a loophole whereby a rogue DNS
4304 administrator could create reverse DNS records to break through a
4305 wildcarded host restriction in an ACL.
4307 37. If a user filter or a system filter that ran in a subprocess used any of
4308 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4309 the wrong values were passed to the pipe command ($thisaddress had the
4310 value of $0, $0 had the value of $1, etc). This bug was introduced by
4311 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4313 38. Improved the line breaking for long SMTP error messages from ACLs.
4314 Previously, if there was no break point between 40 and 75 characters, Exim
4315 left the rest of the message alone. Two changes have been made: (a) I've
4316 reduced the minimum length to 35 characters; (b) if it can't find a break
4317 point between 35 and 75 characters, it looks ahead and uses the first one
4318 that it finds. This may give the occasional overlong line, but at least the
4319 remaining text gets split now.
4321 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4322 file descriptors might be low, and that setting 1000 would always raise it.
4323 It turns out that in some environments, the limit is already over 1000 and
4324 that lowering it causes trouble. So now Exim takes care not to decrease it.
4326 40. When delivering a message, the value of $return_path is set to $sender_
4327 address at the start of routing (routers may change the value). By an
4328 oversight, this default was not being set up when an address was tested by
4329 -bt or -bv, which affected the outcome if any router or filter referred to
4332 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4333 log only when "message" or "log_message" are set. However, if one of the
4334 conditions was an address verification, or a call to a nested ACL, the
4335 messages generated by the underlying test were being passed through. This
4336 no longer happens. The underlying message is available in $acl_verify_
4337 message for both "message" and "log_message" expansions, so it can be
4338 passed through if needed.
4340 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4341 new expansion $bh_ to give the encoded byte string without charset
4342 translation. Translation happens only if iconv() is available; HAVE_ICONV
4343 indicates this at build time. HEADERS_CHARSET gives the charset to
4344 translate to; headers_charset can change it in the configuration, and
4345 "headers charset" can change it in an individual filter file.
4347 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4348 that creates RFC 2047 encoded "words" labels them as that charset instead
4349 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4350 expansion operator; (ii) when Exim creates a From: line for a local
4351 message; (iii) when a header line is rewritten to include a "phrase" part.
4353 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4354 buggy, causing it to skip the first lines of messages whose message ID
4355 ended in 'D'. This would not have bitten before Exim release 4.14, because
4356 message IDs were unlikely to end in 'D' before then. The effect was to have
4357 incorrect size information for certain domains.
4359 45. #include "config.h" was missing at the start of the crypt16.c module. This
4360 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4363 46. If there was a timeout during a "random" callout check, Exim treated it as
4364 a failure of the random address, and carried on sending RSET and the real
4365 address. If the delay was just some slowness somewhere, the response to the
4366 original RCPT would be taken as a response to RSET and so on, causing
4367 mayhem of various kinds.
4369 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4370 when I implemented it. It didn't allow for the fact that some option values
4371 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4372 the right test for positive values.
4374 48. Domain names in DNS records are case-independent. Exim always looks them up
4375 in lower case. Some resolvers return domain names in exactly the case they
4376 appear in the zone file, that is, they may contain uppercase letters. Not
4377 all resolvers do this - some return always lower case. Exim was treating a
4378 change of case by a resolver as a change of domain, similar to a widening
4379 of a domain abbreviation. This triggered its re-routing code and so it was
4380 trying to route what was effectively the same domain again. This normally
4381 caused routing to fail (because the router wouldn't handle the domain
4382 twice). Now Exim checks for this case specially, and just changes the
4383 casing of the domain that it ultimately uses when it transmits the message
4386 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4389 50. If a filter generated a file delivery with a non-absolute name (possible if
4390 no home directory exists for the router), the forbid_file option was not
4393 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4394 the existing equality matching.
4396 52. Exim was using ints instead of ino_t variables in some places where it was
4397 dealing with inode numbers.
4399 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4400 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4401 called TMPDIR, and if it finds it is different, it changes its value.
4403 54. The smtp_printf() function is now made available to local_scan() so
4404 additional output lines can be written before returning. There is also an
4405 smtp_fflush() function to enable the detection of a dropped connection.
4406 The variables smtp_input and smtp_batched_input are exported to
4409 55. Changed the default runtime configuration: the message "Unknown user"
4410 has been removed from the ACL, and instead placed on the localuser router,
4411 using the cannot_route_message feature. This means that any verification
4412 failures that generate their own messages won't get overridden. Similarly,
4413 the "Unrouteable address" message that was in the ACL for unverifiable
4414 relay addresses has also been removed.
4416 56. Added hosts_avoid_esmtp to the smtp transport.
4418 57. The exicyclog script was not checking for the esoteric option
4419 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4420 will work only if exicyclog is run under the appropriate euid.
4422 58. Following a discussion on the list, the rules by which Exim recognises line
4423 endings on incoming messages have been changed. The -dropcr and drop_cr
4424 options are now no-ops, retained only for backwards compatibility. The
4425 following line terminators are recognized: LF CRLF CR. However, special
4426 processing applies to CR:
4428 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4429 nor a local message in the state where . is a terminator.
4431 (ii) If a bare CR is encountered in a header line, an extra space is added
4432 after the line terminator so as not to end the header. The reasoning
4433 behind this is that bare CRs in header lines are most likely either
4434 to be mistakes, or people trying to play silly games.
4436 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4437 was being incorrectly given as 18 bytes larger than it should have been.
4438 This is a VOB (very old bug).
4440 60. This may never have affected anything current, but just in case it has:
4441 When the local host is found other than at the start of a list of hosts,
4442 the local host, those with the same MX, and any that follow, are discarded.
4443 When the list in question was part of a longer list of hosts, the following
4444 hosts (not currently being processed) were also being discarded. This no
4445 longer happens. I'm not sure if this situation could ever has previously
4448 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4451 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4452 header. This is something that is recommended in a new Internet Draft, and
4453 is something that is documented as being done by Sendmail. There are two
4454 possible values. For messages generated by the autoreply transport, Exim
4457 Auto-Submitted: auto-replied
4459 whereas for all other generated messages (e.g. bounces) it adds
4461 Auto-Submitted: auto-generated
4463 63. The "personal" condition in filters now includes a test for the
4464 Auto-Submitted: header. If it contains the string "auto-" the message it
4465 not considered personal.
4467 64. Added rcpt_include_affixes as a generic transport option.
4469 65. Added queue_only_override (default true).
4471 66. Added the syslog_duplication option.
4473 67. If what should have been the first header line of a message consisted of
4474 a space followed by a colon, Exim was mis-interpreting it as a header line.
4475 It isn't of course - it is syntactically invalid and should therefore be
4476 treated as the start of the message body. The misbehaviour could have
4477 caused a number of strange effects, including loss of data in subsequent
4478 header lines, and spool format errors.
4480 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4481 client host had authenticated. This control can now be exercised by an ACL
4482 for more flexibility.
4484 69. By default, callouts do not happen when testing with -bh. There is now a
4485 variant, -bhc, which does actually run the callout code, including
4486 consulting and updating the callout cache.
4488 70. Added support for saslauthd authentication, courtesy of Alexander
4491 71. If statvfs() failed on the spool or log directories while checking their
4492 size for availability, Exim confusingly gave the error "space shortage".
4493 Furthermore, in debugging mode it crashed with a floating point exception.
4494 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4495 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4496 serious problem, Exim now writes to the main and panic logs when this
4497 happens, with details of the failure. It then refuses to accept the
4498 incoming message, giving the message "spool directory problem" or "log
4499 directory problem" with a 421 code for SMTP messages.
4501 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4502 0, 1, and 2 exist, because some OS complain for execs without them (see
4503 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4504 descriptors. However, the code omitted to check that the open succeeded,
4505 causing mysterious errors if for some reason the permissions on /dev/null
4506 got screwed. Now Exim writes a message to the main and panic logs, and
4507 bombs out if it can't open /dev/null.
4509 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4510 interact so that it is all more flexible. It is supposed to remain
4511 backwards compatible. Also added extra_local_interfaces.
4513 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4514 to bomb out with an assertion failure - to the client this appears as a
4515 connection drop. This problem occurs in the part of the code that was taken
4516 from the Samba project. Fortunately, the assertion is in a very simple
4517 function, so I have fixed this by reproducing the function inline in the
4518 one place where it is called, and arranging for authentication to fail
4519 instead of killing the process with assert().
4521 75. The SPA client code was not working when the server requested OEM rather
4522 than Unicode encoding.
4524 76. Added code to make require_files with a specific uid setting more usable in
4525 the case where statting the file as root fails - usually a non-root-mounted
4526 NFS file system. When this happens and the failure is EACCES, Exim now
4527 forks a subprocess and does the per-uid checking as the relevant uid.
4529 77. Added process_log_path.
4531 78. If log_file_path was not explicitly set, a setting of check_log_space or
4532 check_log_inodes was ignored.
4534 79. If a space check for the spool or log partitions fails, the incident is now
4535 logged. Of course, in the latter case the data may get lost...
4537 80. Added the %p formatting code to string_format() so that it can be used to
4538 print addresses in debug_print(). Adjusted all the address printing in the
4539 debugging in store.c to use %p rather than %d.
4541 81. There was a concern that a line of code in smtp_in.c could overflow a
4542 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4543 initially expressed, the concern was not well-founded, because trailing
4544 spaces are removed early. However, if the trailing spaces were followed by
4545 a NULL, they did not get removed, so the overflow was possible. Two fixes
4548 (a) I re-wrote the offending code in a cleaner fashion.
4549 (b) If an incoming SMTP command contains a NULL character, it is rejected
4552 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4553 runs initgroups(), so that if the Exim user is in any additional groups,
4554 they will be used during message reception.
4560 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob