1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.473 2007/02/06 14:49:13 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
63 PH/11 Added control=no_pipelining.
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
72 PH/14 Added log_selector = +pid.
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
76 PH/16 Add ${if forany and ${if forall.
78 PH/17 Added dsn_from option to vary the From: line in DSNs.
80 PH/18 Flush SMTP output before performing a callout, unless control =
81 no_callout_flush is set.
83 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
84 was true (the default) a successful delivery failed to delete the retry
85 item, thus causing premature timeout of the address. The bug is now
88 PH/20 Added hosts_avoid_pipelining to the smtp transport.
94 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
95 fixed by 4.65/MH/01 (is this a record?) are fixed:
97 (i) An empty string was always treated as zero by the numeric comparison
98 operators. This behaviour has been restored.
100 (ii) It is documented that the numeric comparison operators always treat
101 their arguments as decimal numbers. This was broken in that numbers
102 starting with 0 were being interpreted as octal.
104 While fixing these problems I realized that there was another issue that
105 hadn't been noticed. Values of message_size_limit (both the global option
106 and the transport option) were treated as octal if they started with 0.
107 The documentation was vague. These values are now always treated as
108 decimal, and I will make that clear in the documentation.
114 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
115 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
118 MH/01 Don't check that the operands of numeric comparison operators are
119 integers when their expansion is in "skipping" mode (fixes bug
120 introduced by 4.64-PH/07).
122 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
123 child addresses, Exim now panics and dies. Previously, because the count
124 is held in a short int, deliveries were likely to be lost. As such a
125 large number of recipients for a single message is ridiculous
126 (performance will be very, very poor), I have chosen to impose a limit
127 rather than extend the field.
133 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
134 leftover -K file (the existence of which was triggered by #402).
135 While we were at it, introduced process PID as part of the -K
136 filename. This should rule out race conditions when creating
139 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
140 processing considerably. Previous code took too long for large mails,
141 triggering a timeout which in turn triggers #401.
143 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
144 in the DK code in transports.c. sendfile() is not really portable,
145 hence the _LINUX specificness.
147 TF/01 In the add_headers option to the mail command in an Exim filter,
148 there was a bug that Exim would claim a syntax error in any
149 header after the first one which had an odd number of characters
152 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
153 callout verification, Exim cached a "reject" for the entire domain. This
154 is correct for most verifications, but it is not correct for a recipient
155 verification with use_sender or use_postmaster set, because in that case
156 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
157 case of MAIL FROM:<> rejection from other early rejections (e.g.
158 rejection of HELO). When verifying a recipient using a non-null MAIL
159 address, the cache is ignored if it shows MAIL FROM:<> rejection.
160 Whatever the result of the callout, the value of the domain cache is
161 left unchanged (for any other kind of callout, getting as far as trying
162 RCPT means that the domain itself is ok).
164 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
167 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
168 manifest itself as EPIPE rather than ECONNECT. When tidying away a
169 session, the daemon ignores ECONNECT errors and logs others; it now
170 ignores EPIPE as well.
172 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
173 (quoted-printable decoding).
175 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
176 later the small subsequent patch to fix an introduced bug.
178 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
180 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
182 PH/08 An error is now given if message_size_limit is specified negative.
184 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
185 to be given (somewhat) arbitrary names.
187 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
190 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
191 miscellaneous code fixes
193 PH/10 Added the log_reject_target ACL modifier to specify where to log
196 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
197 hostname. This is wrong, because it relates to the incoming message (and
198 probably the interface on which it is arriving) and not to the outgoing
199 callout (which could be using a different interface). This has been
200 changed to use the value of the helo_data option from the smtp transport
201 instead - this is what is used when a message is actually being sent. If
202 there is no remote transport (possible with a router that sets up host
203 addresses), $smtp_active_hostname is used.
205 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
206 tweaks were necessary in order to get it to work (see also 21 below):
207 (a) The code assumed that strncpy() returns a negative number on buffer
208 overflow, which isn't the case. Replaced with Exim's string_format()
210 (b) There were several signed/unsigned issues. I just did the minimum
211 hacking in of casts. There is scope for a larger refactoring.
212 (c) The code used strcasecmp() which is not a standard C function.
213 Replaced with Exim's strcmpic() function.
214 (d) The code set only $1; it now sets $auth1 as well.
215 (e) A simple test gave the error "authentication client didn't specify
216 service in request". It would seem that Dovecot has changed its
217 interface. Fortunately there's a specification; I followed it and
218 changed what the client sends and it appears to be working now.
220 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
223 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
224 address A is aliased to B and C, where B exists and C does not. Without
225 -v the output is "A verified" because verification stops after a
226 successful redirection if more than one address is generated. However,
227 with -v the child addresses are also verified. Exim was outputting "A
228 failed to verify" and then showing the successful verification for C,
229 with its parentage. It now outputs "B failed to verify", showing B's
230 parentage before showing the successful verification of C.
232 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
233 look up a TXT record in a specific list after matching in a combined
236 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
237 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
238 they consult the DNS. I had assumed they would set it the way they
239 wanted; and indeed my experiments on Linux seem to show that in some
240 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
241 To be on the safe side, however, I have now made the interface to
242 host_find_byname() similar to host_find_bydns(), with an argument
243 containing the DNS resolver options. The host_find_byname() function now
244 sets these options at its start, just as host_find_bydns() does. The smtp
245 transport options dns_qualify_single and dns_search_parents are passed to
246 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
247 of host_find_byname() use the default settings of RES_DEFNAMES
248 (qualify_single) but not RES_DNSRCH (search_parents).
250 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
251 spool_read_header() do less string testing, by means of a preliminary
252 switch on the second character of optional "-foo" lines. (This is
253 overdue, caused by the large number of possibilities that now exist.
254 Originally there were few.) While I was there, I also converted the
255 str(n)cmp tests so they don't re-test the leading "-" and the first
256 character, in the hope this might squeeze out yet more improvement.
258 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
259 flag allowing group syntax was set by the header_syntax check but not
260 turned off, possible causing trouble later; (2) The flag was not being
261 set at all for the header_verify test, causing "group"-style headers to
262 be rejected. I have now set it in this case, and also caused header_
263 verify to ignore an empty address taken from a group. While doing this, I
264 came across some other cases where the code for allowing group syntax
265 while scanning a header line wasn't quite right (mostly, not resetting
266 the flag correctly in the right place). These bugs could have caused
267 trouble for malformed header lines. I hope it is now all correct.
269 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
270 with the "reply" argument non-NULL. The code, however (which originally
271 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
272 but it didn't always do it. This confused somebody who was copying the
273 code for some other use. I have removed all the tests.
275 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
276 feature that was used to support insecure browsers during the U.S. crypto
277 embargo. It requires special client support, and Exim is probably the
278 only MTA that supported it -- and would never use it because real RSA is
279 always available. This code has been removed, because it had the bad
280 effect of slowing Exim down by computing (never used) parameters for the
281 RSA_EXPORT functionality.
283 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
284 authenticator to fail if there's a tab character in the incoming data
285 (there should never be unless someone is messing about, as it's supposed
286 to be base64-encoded). Also added, on Timo's advice, the "secured" option
287 if the connection is using TLS or if the remote IP is the same as the
288 local IP, and the "valid-client-cert option" if a client certificate has
291 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
292 authenticators. This can be used for authorization after authentication
293 succeeds. (In the case of plaintext, it servers for both authentication
296 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
297 if any retry times were supplied.
299 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
300 connection, where there is no client IP address to check. In this
301 situation, the verify now always succeeds.
303 PH/25 Applied John Jetmore's -Mset patch.
305 PH/26 Added -bem to be like -Mset, but loading a message from a file.
307 PH/27 In a string expansion for a processed (not raw) header when multiple
308 headers of the same name were present, leading whitespace was being
309 removed from all of them, but trailing whitespace was being removed only
310 from the last one. Now trailing whitespace is removed from each header
311 before concatenation. Completely empty headers in a concatenation (as
314 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
315 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
317 PH/29 [Removed. This was a change that I later backed out, and forgot to
318 correct the ChangeLog entry (that I had efficiently created) before
319 committing the later change.]
321 PH/30 Exim was sometimes attempting to deliver messages that had suffered
322 address errors (4xx response to RCPT) over the same connection as other
323 messages routed to the same hosts. Such deliveries are always "forced",
324 so retry times are not inspected. This resulted in far too many retries
325 for the affected addresses. The effect occurred only when there were more
326 hosts than the hosts_max_try setting in the smtp transport when it had
327 the 4xx errors. Those hosts that it had tried were not added to the list
328 of hosts for which the message was waiting, so if all were tried, there
329 was no problem. Two fixes have been applied:
331 (i) If there are any address or message errors in an SMTP delivery, none
332 of the hosts (tried or untried) are now added to the list of hosts
333 for which the message is waiting, so the message should not be a
334 candidate for sending over the same connection that was used for a
335 successful delivery of some other message. This seems entirely
336 reasonable: after all the message is NOT "waiting for some host".
337 This is so "obvious" that I'm not sure why it wasn't done
338 previously. Hope I haven't missed anything, but it can't do any
339 harm, as the worst effect is to miss an optimization.
341 (ii) If, despite (i), such a delivery is accidentally attempted, the
342 routing retry time is respected, so at least it doesn't keep
343 hammering the server.
345 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
346 in ${readsocket because some servers need this prod.
348 PH/32 Added some extra debug output when updating a wait-xxx database.
350 PH/33 The hint "could be header name not terminated by colon", which has been
351 given for certain expansion errors for a long time, was not being given
352 for the ${if def:h_colon_omitted{... case.
354 PH/34 The spec says: "With one important exception, whenever a domain list is
355 being scanned, $domain contains the subject domain." There was at least
356 one case where this was not true.
358 PH/35 The error "getsockname() failed: connection reset by peer" was being
359 written to the panic log as well as the main log, but it isn't really
360 panic-worthy as it just means the connection died rather early on. I have
361 removed the panic log writing for the ECONNRESET error when getsockname()
364 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
365 runs only) independently of the message's sender address. This meant
366 that, if the 4xx error was in fact related to the sender, a different
367 message to the same recipient with a different sender could confuse
368 things. In particualar, this can happen when sending to a greylisting
369 server, but other circumstances could also provoke similar problems.
370 I have changed the default so that the retry time for these errors is now
371 based a combination of the sender and recipient addresses. This change
372 can be overridden by setting address_retry_include_sender=false in the
375 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
376 remote server are returned as part of bounce messages. This was not
377 happening for LMTP over a pipe (the lmtp transport), but now it is the
378 same for both kinds of LMTP.
380 PH/38 Despite being documented as not happening, Exim was rewriting addresses
381 in header lines that were in fact CNAMEs. This is no longer the case.
383 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
384 and queue runs started by the daemon processed all messages. This has
385 been fixed so that -R and -S can now usefully be given with -q<time>.
387 PH/40 Import PCRE release 6.7 (fixes some bugs).
389 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
391 PH/42 Give an error if -q is specified more than once.
393 PH/43 Renamed the variables $interface_address and $interface_port as
394 $received_ip_address and $received_port, to make it clear that these
395 values apply to message reception, and not to the outgoing interface when
396 a message is delivered. (The old names remain recognized, of course.)
398 PH/44 There was no timeout on the connect() call when using a Unix domain
399 socket in the ${readsocket expansion. There now is.
401 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
402 be meaningful with "accept".
404 SC/01 Eximstats V1.43
405 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
407 SC/02 Eximstats V1.44
408 Use a glob alias rather than an array ref in the generated
409 parser. This improves both readability and performance.
411 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
412 Collect SpamAssassin and rejection statistics.
413 Don't display local sender or destination tables unless
414 there is data to show.
415 Added average volumes into the top table text output.
417 SC/04 Eximstats V1.46
418 Collect data on the number of addresses (recipients)
419 as well as the number of messages.
421 SC/05 Eximstats V1.47
422 Added 'Message too big' to the list of mail rejection
423 reasons (thanks to Marco Gaiarin).
425 SC/06 Eximstats V1.48
426 Mainlog lines which have GMT offsets and are too short to
427 have a flag are now skipped.
429 SC/07 Eximstats V1.49 (Alain Williams)
430 Added the -emptyok flag.
432 SC/08 Eximstats V1.50
433 Fixes for obtaining the IP address from reject messages.
435 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
436 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
437 whitesspace changes from 4.64-PH/27
439 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
442 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
443 are found, allow negative numbers in numeric criteria)
445 JJ/06 exipick.20061117.2, added new $message_body_missing variable
447 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
448 to match changes made in 4.64-PH/43
450 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
452 PH/47 Put in an explicit test for a DNS lookup of an address record where the
453 "domain" is actually an IP address, and force a failure. This locks out
454 those revolvers/nameservers that support "A-for-A" lookups, in
455 contravention of the specifications.
457 PH/48 When a host name was looked up from an IP address, and the subsequent
458 forward lookup of the name timed out, the host name was left in
459 $sender_host_name, contrary to the specification.
461 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
462 restricted to single-key lookups, Exim was not diagnosing an error if
463 * or *@ was used with a query-style lookup.
465 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
467 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
468 long ago, but noone interested enough thought of it. Let's just say that
469 the "1.1" means that there are some new functions that weren't there at
470 some point in the past.
472 PH/51 Error processing for expansion failure of helo_data from an smtp
473 transport during callout processing was broken.
475 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
476 tested/used via the -bh/-bhc/-bs options.
478 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
479 bug, fixed in subsequent PCRE releases).
481 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
482 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
484 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
490 SC/01 Use a glob alias rather than an array ref in eximstats generated
491 parser. This improves both readability and performance.
493 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
494 Don't display local sender or destination tables in eximstats unless
495 there is data to show.
496 Added average volumes into the eximstats top table text output.
498 SC/03 Collect data on the number of addresses (recipients) as well
499 as the number of messages in eximstats.
501 TF/01 Correct an error in the documentation for the redirect router. Exim
502 does (usually) call initgroups() when daemonizing.
504 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
505 with consistent privilege compared to when running as a daemon.
507 TF/03 Note in the spec that $authenticated_id is not set for local
508 submissions from trusted users.
510 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
511 Thanks to Dean Brooks <dean@iglou.com> for the patch.
513 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
514 by adding some example configuration directives to the default
515 configuration file. A little bit of work is required to uncomment the
516 directives and define how usernames and passwords are checked, but
517 there is now a framework to start from.
519 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
520 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
521 without this. I don't know how relevant this is to other LDAP libraries.
523 PH/02 Add the verb name to the "unknown ACL verb" error.
525 PH/03 Magnus Holmgren's patch for filter_prepend_home.
527 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
529 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
530 directory not expanded when it should be if an expanded home directory
531 was set for the address (which is overridden by the transport).
533 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
536 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
537 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
538 because it is too late at that time, and has no effect.
540 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
541 security issue with \' (bugzilla #107). I could not use the
542 PQescapeStringConn() function, because it needs a PGconn value as one of
545 PH/08 When testing addresses using -bt, indicate those final addresses that
546 are duplicates that would not cause an additional delivery. At least one
547 person was confused, thinking that -bt output corresponded to deliveries.
548 (Suppressing duplicates isn't a good idea as you lose the information
549 about possibly different redirections that led to the duplicates.)
551 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
552 systems where poll() doesn't work, in particular OS X.
554 PH/10 Added more information to debugging output for retry time not reached.
556 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
557 operations in malware.c.
559 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
562 PH/13 If write_rejectlog was set false when logging was sent to syslog with
563 syslog_duplication set false, log lines that would normally be written
564 both the the main log and to the reject log were not written to syslog at
567 PH/14 In the default configuration, change the use of "message" in ACL warn
568 statements to "add_header".
570 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
571 not followed by a command (e.g. "seen endif").
573 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
574 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
577 PH/17 Added extra conditions to the default value of delay_warning_condition
581 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
582 { match{$h_precedence:}{(?i)bulk|list|junk} } \
583 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
586 The Auto-Submitted: and various List- headers are standardised, whereas I
587 don't think Precedence: ever was.
589 PH/18 Refactored debugging code in route_finduser() to show more information,
590 in particular, the error code if getpwnam() issues one.
592 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
593 This is apparently needed in addition to the PH/07 change above to avoid
594 any possible encoding problems.
596 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
597 but not after initializing Perl.
599 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
600 output them only if debugging. By default they are written stderr,
601 apparently, which is not desirable.
603 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
606 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
609 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
611 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
612 authenticated or an ident call has been made. Suppress the default
613 values for $authenticated_id and $authenticated_sender (but permit -oMai
614 and -oMas) when testing with -bh.
616 PH/24 Re-jigged the order of the tests in the default configuration so that the
617 tests for valid domains and recipients precede the DNS black list and CSA
618 tests, on the grounds that those ones are more expensive.
620 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
621 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
622 This bug exists in every version of Exim that I still have, right back to
625 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
626 However, an attempt to turn on case-sensitivity in a regex key by
627 including (?-i) didn't work because the subject string was already
628 lowercased, and the effects were non-intuitive. It turns out that a
629 one-line patch can be used to allow (?-i) to work as expected.
635 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
636 other effects) broke the use of negated acl sub-conditions.
638 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
641 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
642 "Deny" causes Exim to reject the incoming connection with a 554 error.
643 Unfortunately, if there is a major crisis, such as a disk failure,
644 tcp-wrappers gives "deny", whereas what one would like would be some
645 kind of temporary error. A kludge has been added to help with this.
646 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
647 554 error is used if errno is still zero or contains ENOENT (which occurs
648 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
651 PH/03 Add -lutil to the default FreeBSD LIBS setting.
653 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
654 errors. Otherwise a message that provokes a temporary error (when other
655 messages do not) can cause a whole host to time out.
657 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
658 addresses were routed directly to files or pipes from a redirect router.
659 File deliveries just didn't batch; pipe deliveries might have suffered
662 PH/06 A failure to get a lock for a hints database would erroneously always say
663 "Failed to get write lock", even when it was really a read lock.
665 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
666 of 0600. This has been changed to use the value of the lockfile_mode
667 option (which defaults to 0600).
669 PH/08 Applied small patch from the Sieve maintainer.
671 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
672 folder from quota calculations, a direct delivery into this folder messed
673 up the contents of the maildirsize file. This was because the regex was
674 used only to exclude .Trash (or whatever) when the size of the mailbox
675 was calculated. There was no check that a delivery was happening into an
676 excluded directory. This bug has been fixed by ignoring all quota
677 processing for deliveries into excluded directories.
679 PH/10 Added the maildirfolder_create_regex option to appendfile.
685 PH/01 The code for finding all the local interface addresses on a FreeBSD
686 system running IPv6 was broken. This may well have applied to all BSD
687 systems, as well as to others that have similar system calls. The broken
688 code found IPv4 interfaces correctly, but gave incorrect values for the
689 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
690 that it would not match correctly against @[] and not recognize the IPv6
693 PH/02 The ipliteral router was not recognizing addresses of the form user@
694 [ipv6:....] because it didn't know about the "ipv6:" prefix.
696 PH/03 Added disable_ipv6.
698 PH/04 Changed $reply_address to use the raw form of the headers instead of the
699 decoded form, because it is most often used to construct To: headers
700 lines in autoreplies, and the decoded form may well be syntactically
701 invalid. However, $reply_address has leading white space removed, and all
702 newlines turned into spaces so that the autoreply transport does not
705 PH/05 If group was specified without a user on a router, and no group or user
706 was specified on a transport, the group from the router was ignored.
708 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
709 for visible compile-time settings that can be used to change these
710 numbers, for those that want even more. Backwards compatibility with old
711 spool files has been maintained. However, going back to a previous Exim
712 release will lost any variables that are in spool files.
714 PH/07 Two small changes when running in the test harness: increase delay when
715 passing a TCP/IP connection to a new process, in case the original
716 process has to generate a bounce, and remove special handling of
717 127.0.0.2 (sic), which is no longer necessary.
719 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
720 be the same on different OS.
722 PH/09 Moved a debug statement in filter processing to avoid a race problem when
725 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
726 whether --show-vars was specified or not
728 JJ/02 exipick: Added support for new ACL variable spool format introduced
731 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
732 syntactically invalid From: or Reply-to: line, and a filter used this to
733 generate an autoreply, and therefore failed to obtain an address for the
734 autoreply, Exim could try to deliver to a non-existent relative file
735 name, causing unrelated and misleading errors. What now happens is that
736 it logs this as a hard delivery error, but does not attempt to create a
739 PH/11 The exinext utility has a -C option for testing purposes, but although
740 the given file was scanned by exinext itself; it wasn't being passed on
741 when Exim was called.
743 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
744 an end-of-file indication when reading a command response.
746 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
747 compiled. In many other places in Exim, IPv6 addresses are always
748 recognized, so I have changed this. It also means that IPv4 domain
749 literals of the form [IPV4:n.n.n.n] are now always recognized.
751 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
752 used if the router is not running as root, for example, when verifying at
753 ACL time, or when using -bh. The debugging output from this situation was
754 non-existent - all you got was a failure to exec. I have made two
757 (a) Failures to set uid/gid, the current directory, or a process leader
758 in a subprocess such as that created by queryprogram now generate
759 suitable debugging ouput when -d is set.
761 (b) The queryprogram router detects when it is not running as root,
762 outputs suitable debugging information if -d is set, and then runs
763 the subprocess without attempting to change uid/gid.
765 PH/15 Minor change to Makefile for building test_host (undocumented testing
768 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
769 additional section of a DNS packet that returns MX or SRV records.
770 Instead, it always explicitly searches for A/AAAA records. This avoids
771 major problems that occur when a DNS server includes only records of one
772 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
773 fixed another bug: if SRV records were looked up and the corresponding
774 address records were *not* found in the additional section, the port
775 values from the SRV records were lost.
777 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
778 using the correct key (the original address) when searching the retry
779 rules in order to find which one to use for generating the retry hint.
781 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
782 adding the default one. Similarly, if it contains a Reply-To: header, the
783 errors_reply_to option, if set, is not used.
785 PH/19 When calculating a retry time, Exim used to measure the "time since
786 failure" by looking at the "first failed" field in the retry record. Now
787 it does not use this if it is later than than the arrival time of the
788 message. Instead it uses the arrival time. This makes for better
789 behaviour in cases where some deliveries succeed, thus re-setting the
790 "first failed" field. An example is a quota failure for a huge message
791 when small messages continue to be delivered. Without this change, the
792 "time since failure" will always be short, possible causing more frequent
793 delivery attempts for the huge message than are intended.
794 [Note: This change was subsequently modified - see PH/04 for 4.62.]
796 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
797 $1, $2, $3) because the numerical variables can be reset during some
798 expansion items (e.g. "match"), thereby losing the authentication data.
800 PH/21 Make -bV show the size of off_t variables so that the test suite can
801 decide whether to run tests for quotas > 2G.
803 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
804 mailbox_size, and mailbox_filecount in the appendfile transport. If a
805 filecount value is greater than 2G or if a quota value is greater than 2G
806 on a system where the size of off_t is not greater than 4, a panic error
809 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
810 never match. The debug and -bh output now contains an explicit error
811 message indicating a malformed IPv4 address or mask.
813 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
814 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
817 PH/25 Do not write to syslog when running in the test harness. The only
818 occasion when this arises is a failure to open the main or panic logs
819 (for which there is an explicit test).
821 PH/26 Added the /no_tell option to "control=freeze".
823 PH/27 If a host name lookup failed very early in a connection, for example, if
824 the IP address matched host_lookup and the reverse lookup yielded a name
825 that did not have a forward lookup, an error message of the form "no IP
826 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
827 could be logged. Now it outputs the IP address instead of "NULL".
829 PH/28 An enabling patch from MH: add new function child_open_exim2() which
830 allows the sender and the authenticated sender to be set when
831 submitting a message from within Exim. Since child_open_exim() is
832 documented for local_scan(), the new function should be too.
834 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
835 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
836 results in an empty string is now treated as unset.
838 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
840 PH/31 Added sender_verify_fail logging option.
842 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
843 needed by RFC 822 but not by RFC 2822 was commented out. I have now
844 tidied the source and removed it altogether.
846 PH/33 When a queue run was abandoned because the load average was too high, a
847 log line was always written; now it is written only if the queue_run log
848 selector is set. In addition, the log line for abandonment now contains
849 information about the queue run such as the pid. This is always present
850 in "start" and "stop" lines but was omitted from the "abandon" line.
852 PH/34 Omit spaces between a header name and the colon in the error message that
853 is given when verify = headers_syntax fails (if there are lots of them,
854 the message gets confusing).
856 PH/35 Change the default for dns_check_names_pattern to allow slashes within
857 names, as there are now some PTR records that contain slashes. This check
858 is only to protect against broken name servers that fall over on strange
859 characters, so the fact that it applies to all lookups doesn't matter.
861 PH/36 Now that the new test suite is complete, we can remove some of the
862 special code in Exim that was needed for the old test suite. For example,
863 sorting DNS records because real resolvers return them in an arbitrary
864 order. The new test suite's fake resolver always returns records in the
867 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
868 bounces) except when queue_only is set, to avoid logging races between
869 the different processes.
871 PH/38 Panic-die if .include specifies a non-absolute path.
873 PH/39 A tweak to the "H" retry rule from its user.
875 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
876 a label. They prevented compilation on older perls.
878 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
879 a warning to be raised on newish perls.
881 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
882 on queue. Changes to match documented behaviour of showing count of
883 messages matching specified criteria.
885 PH/40 Changed the default ident timeout from 30s to 5s.
887 PH/41 Added support for the use of login_cap features, on those BSD systems
888 that have them, for controlling the resources used by pipe deliveries.
890 PH/42 The content-scanning code uses fopen() to create files in which to put
891 message data. Previously it was not paying any attention to the mode of
892 the files. Exim runs with umask(0) because the rest of the code creates
893 files with open(), and sets the required mode explicitly. Thus, these
894 files were ending up world-writeable. This was not a big issue, because,
895 being within the spool directory, they were not world-accessible. I have
896 created a function called modefopen, which takes an additional mode
897 argument. It sets umask(777), creates the file, chmods it to the required
898 mode, then resets the umask. All the relevant calls to fopen() in the
899 content scanning code have been changed to use this function.
901 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
902 to 24 hours. This avoids potential overflow problems when processing G
903 and H retry rules. I suspect nobody ever tinkers with this value.
905 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
907 PH/45 When the plaintext authenticator is running as a client, the server's
908 challenges are checked to ensure they are valid base64 strings. By
909 default, the authentication attempt is cancelled if an invalid string is
910 received. Setting client_ignore_invalid_base64 true ignores these errors.
911 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
912 they are received. Thus, the responses can be made to depend on the
913 challenges. If an invalid string is ignored, an empty string is placed in
916 PH/46 Messages that are created by the autoreply transport now contains a
917 References: header, in accordance with RFCs 2822 and 3834.
919 PH/47 Added authenticated_sender_force to the smtp transport.
921 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
923 PH/49 Installed latest patch from the Sieve maintainer.
925 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
926 was also set, the appendfile transport was unnecessarily scanning a
927 directory of message files (e.g. for maildir delivery) to find the count
928 of files (along with the size), even though it did not need this
929 information. It now does the scan only if it needs to find either the
930 size of the count of files.
932 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
934 PH/52 Two bugs concerned with error handling when the smtp transport is
937 (i) Exim was not creating retry information for temporary errors given
938 for individual recipients after the DATA command when the smtp transport
939 was used in LMTP mode. This meant that they could be retried too
940 frequently, and not timed out correctly.
942 (ii) Exim was setting the flag that allows error details to be returned
943 for LMTP errors on RCPT commands, but not for LMTP errors for individual
944 recipients that were returned after the DATA command.
946 PH/53 This is related to PH/52, but is more general: for any failing address,
947 when detailed error information was permitted to be returned to the
948 sender, but the error was temporary, then after the final timeout, only
949 "retry timeout exceeded" was returned. Now it returns the full error as
950 well as "retry timeout exceeded".
952 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
953 do this, and (what is worse) MTAs that accept it.
955 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
956 will now be deprecated.
958 PH/56 New os.c-cygwin from the Cygwin maintainer.
960 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
961 formats (previously only available in exim formats via -bpr, -bpru,
962 and -bpra. Now also available in native and exiqgrep formats)
964 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
965 with very large, slow to parse queues
967 JJ/08 exipick: added ! as generic prefix to negate any criteria format
969 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
971 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
972 responses to authentication challenges, though it was showing the
973 challenges; (ii) I've removed the CR characters from the debug output for
976 PH/58 Allow for the insertion of a newline as well as a space when a string
977 is turned into more than one encoded-word during RFC 2047 encoding. The
978 Sieve code now uses this.
980 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
981 data_4xx, lost_connection, tls_required.
983 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
984 message was being sent as an SMTP response.
986 PH/61 Add -l and -k options to exicyclog.
988 PH/62 When verifying, if an address was redirected to one new address, so that
989 verification continued, and the new address failed or deferred after
990 having set something in $address_data, the value of $address_data was not
991 passed back to the ACL. This was different to the case when no
992 redirection occurred. The value is now passed back in both cases.
994 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
995 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
996 use login_cap.h, so on its own it isn't the distinguishing feature. The
997 new name refers directly to the setclassresources() function.
999 PH/65 Added configuration files for NetBSD3.
1001 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1003 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1004 is preferred over IPv4.
1006 PH/68 The bounce_return_message and bounce_return_body options were not being
1007 honoured for bounces generated during the reception of non-SMTP messages.
1008 In particular, this applied to messages rejected by the ACL. This bug has
1009 been fixed. However, if bounce_return_message is true and bounce_return_
1010 body is false, the headers that are returned for a non-SMTP message
1011 include only those that have been read before the error was detected.
1012 (In the case of an ACL rejection, they have all been read.)
1014 PH/69 The HTML version of the specification is now built in a directory called
1015 spec_html instead of spec.html, because the latter looks like a path with
1016 a MIME-type, and this confuses some software.
1018 PH/70 Catch two compiler warnings in sieve.c.
1020 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1021 function verify_get_ident() calls ip_connect() to connect a socket, but
1022 if the "connect()" function timed out, ip_connect() used to close the
1023 socket. However, verify_get_ident() also closes the socket later, and in
1024 between Exim writes to the log, which may get opened at this point. When
1025 the socket was closed in ip_connect(), the log could get the same file
1026 descriptor number as the socket. This naturally causes chaos. The fix is
1027 not to close the socket in ip_connect(); the socket should be closed by
1028 the function that creates it. There was only one place in the code where
1029 this was missing, in the iplookup router, which I don't think anybody now
1030 uses, but I've fixed it anyway.
1032 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1033 well as to direct DNS lookups. Otherwise the handling of names in host
1034 lists is inconsistent and therefore confusing.
1040 PH/01 Two changes to the default runtime configuration:
1042 (1) Move the checks for relay_from_hosts and authenticated clients from
1043 after to before the (commented out) DNS black list checks.
1045 (2) Add control=submission to the relay_from_hosts and authenticated
1046 clients checks, on the grounds that messages accepted by these
1047 statements are most likely to be submissions.
1049 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1051 (1) Generate an error if the third argument for the ${prvs expansion is
1054 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1057 (3) Reset the variables that are obtained from the first argument of
1058 ${prvscheck and used in the second argument before leaving the code,
1059 because their memory is reclaimed, so using them afterwards may do
1062 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1063 one (it's much easier than Tom thought :-).
1065 (5) Because of (4), we can now allow for the use of $prvscheck_result
1066 inside the third argument.
1068 PH/03 For some reason, the default setting of PATH when running a command from
1069 a pipe transport was just "/usr/bin". I have changed it to
1072 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1073 anything to be listed in the output from -bV.
1075 PH/05 When a filter generated an autoreply, the entire To: header line was
1076 quoted in the delivery log line, like this:
1078 => >A.N.Other <ano@some.domain> <original@ddress> ...
1080 This has been changed so that it extracts the operative address. There
1081 may be more than one such address. If so, they are comma-separated, like
1084 => >ano@some.domain,ona@other.domain <original@ddress> ...
1086 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1087 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1088 not being looked up in the rDNS to get a host name, Exim was showing the
1089 IP address twice in Received: lines, even though the IP addresses were
1090 identical. For example:
1092 Received: from [1.2.3.4] (helo=[1.2.3.4])
1094 However, if the real host name was known, it was omitting the HELO data
1095 if it matched the actual IP address. This has been tidied up so that it
1096 doesn't show the same IP address twice.
1098 PH/07 When both +timestamp and +memory debugging was on, the value given by
1099 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1100 called by the debug printing, thereby overwriting the timestamp buffer.
1101 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1103 PH/08 When the original message was included in an autoreply transport, it
1104 always said "this is a copy of the message, including all the headers",
1105 even if body_only or headers_only was set. It now gives an appropriate
1108 PH/09 Applied a patch from the Sieve maintainer which:
1110 o fixes some comments
1111 o adds the (disabled) notify extension core
1112 o adds some debug output for the result of if/elsif tests
1113 o points to the current vacation draft in the documentation
1114 and documents the missing references header update
1118 o fixes a bug in processing the envelope test (when testing
1119 multiple envelope elements, the last element determinted the
1122 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1123 Electronic Mail") by including:
1125 Auto-submitted: auto-generated
1127 in the messages that it generates (bounce messages and others, such as
1128 warnings). In the case of bounce messages for non-SMTP mesages, there was
1129 also a typo: it was using "Auto_submitted" (underscore instead of
1130 hyphen). Since every message generated by Exim is necessarily in response
1131 to another message, thes have all been changed to:
1133 Auto-Submitted: auto-replied
1135 in accordance with these statements in the RFC:
1137 The auto-replied keyword:
1139 - SHOULD be used on messages sent in direct response to another
1140 message by an automatic process,
1142 - MUST NOT be used on manually-generated messages,
1144 - MAY be used on Delivery Status Notifications (DSNs) and Message
1145 Disposition Notifications (MDNs),
1147 - MUST NOT be used on messages generated by automatic or periodic
1148 processes, except for messages which are automatic responses to
1151 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1152 to the default Received: header definition.
1154 PH/12 Added log selector acl_warn_skipped (default on).
1156 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1157 variables because (a) they are in the wrong storage pool and (b) even if
1158 they were copied, it wouldn't work properly because of the caching.
1160 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1161 checking when decoding. Apparently there are clients that generate
1162 overlong encoded strings. Why am I not surprised?
1164 PH/15 If the first argument of "${if match_address" was not empty, but did not
1165 contain an "@" character, Exim crashed. Now it writes a panic log message
1166 and treats the condition as false.
1168 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1170 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1171 extension "envelope-auth". The code is finished and in agreement with
1172 other implementations, but there is no documentation so far and in fact,
1173 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1174 not changing the active code.
1176 Print executed "if" and "elsif" statements when debugging is used. This
1177 helps a great deal to understand what a filter does.
1179 Document more things not specified clearly in RFC3028. I had all this
1180 sorted out, when out of a sudden new issues came to my mind. Oops."
1182 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1185 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1186 canonical form (as documented). However, after a host name lookup from
1187 the IP address, check_host() was doing a simple string comparison with
1188 addresses acquired from the DNS when checking that the found name did
1189 have the original IP as one of its addresses. Since any found IPv6
1190 addresses are likely to be in abbreviated form, the comparison could
1191 fail. Luckily, there already exists a function for doing the comparison
1192 by converting both addresses to binary, so now that is used instead of
1193 the text comparison.
1195 PH/20 There was another similar case to PH/19, when a complete host name was
1196 given in a host list; looking up its IP address could give an abbreviated
1197 form, whereas the current host's name might or might not be abbreviated.
1198 The same fix has been applied.
1204 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1205 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1208 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1209 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1211 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1213 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1214 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1215 Also, for "Auto-Submitted", treat anything other than "no" as
1216 non-personal, in accordance with RFC 3834. (Previously it treated
1217 anything starting "auto-" as non-personal.)
1219 TF/01 The control=submission/name=... option had a problem with syntax
1220 errors if the name included a slash character. The /name= option
1221 now slurps the rest of the string, so it can include any characters
1222 but it must come last in the list of options (after /sender_retain
1225 PH/05 Some modifications to the interface to the fake nameserver for the new
1233 TK/01 Added the "success_on_redirect" address verification option. See
1234 NewStuff for rationale and an example.
1236 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1238 PH/02 Patch to exigrep to allow it to work on syslog lines.
1240 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1241 fread() to skip over the body file's header line, because in Cygwin the
1242 header line is locked and is inaccessible.
1244 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1245 co-exist for some time) to make it clear that it is the Exim ID that is
1246 referenced, not the Message-ID: header line.
1248 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1249 string_format() function, because snprintf() does not exist on all
1252 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1253 ${stat: expansion item.
1255 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1256 protocol synchronization error", to keep the pedants happy.
1258 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1259 well as for IRIX systems, when gcc is being used. See the host.c source
1262 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1264 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1267 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1268 transport and to the smtp transport in LMTP mode.
1270 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1272 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1273 run a filter in a subprocess. This could lead to confusion in subsequent
1274 lookups in the parent process. There should also be a search_tidyup() at
1275 the end of the subprocess.
1277 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1278 only if the host matched helo_try_verify_hosts, which caused the
1279 verification to occur when the EHLO/HELO command was issued. The ACL just
1280 tested the remembered result. Now, if a previous verification attempt has
1281 not happened, "verify = helo" does it there and then.
1283 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1285 TK/03 Fix log output including CR from clamd.
1287 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1288 exist provoked a memory error which could cause a segfault.
1290 PH/15 Installed PCRE 6.2
1292 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1294 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1295 of the problem. Specifically, suggested +O2 rather than +O1 for the
1298 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1300 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1301 record, the port was not correctly logged when the outgoing_port log
1302 selector was set (it logged the transort's default port).
1304 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1305 fallback_hosts, and "hosts" in the smtp transport.
1307 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1308 host errors such as "Connection refused".
1310 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1311 authentication with radiusclient 0.4.9:
1313 - Error returned from rc_read_config was caught wrongly
1314 - Username/password not passed on to radius server due to wrong length.
1316 The presumption is that some radiusclient API changes for 4.51/PH/17
1317 were not taken care of correctly. The code is still untested by me (my
1318 Linux distribution still has 0.3.2 of radiusclient), but it was
1319 contributed by a Radius user.
1321 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1322 expanding the "port" option of the smtp transport.
1324 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1325 while reading a MIME header. Thanks to Tom Hughes for a patch.
1327 PH/24 Include config.h inside local_scan.h so that configuration settings are
1330 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1331 that in an ACL for RCPT (for example), you can examine exactly what was
1334 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1335 commands, but it was not correctly comparing the address with the actual
1336 client host address. Thus, it would show the EHLO address in Received:
1337 header lines when this was not necessary.
1339 PH/27 Added the % operator to ${eval:}.
1341 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1342 it should be ignoring failures (because with -C, for example, it has lost
1343 privilege). It wasn't ignoring creation failures other than "already
1346 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1349 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1350 because an input line was too long, either on its own, or by virtue of
1351 too many continuations, the temporary file was not being removed, and the
1352 return code was incorrect.
1354 PH/31 Missing "BOOL" in function definition in filtertest.c.
1356 PH/32 Applied Sieve patches from the maintainer.
1358 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1360 PH/33 Added "verify = not_blind".
1362 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1363 Local/Makefile (with some defaults set). These are used in built scripts
1364 such as exicyclog, but they have never been used in the exim_install
1365 script (though there are many overriding facilities there). I have
1366 arranged that the exim_install script now takes note of these two
1369 PH/35 Installed configuration files for Dragonfly.
1371 PH/36 When a locally submitted message by a trusted user did not contain a
1372 From: header, and the sender address was obtained from -f or from an SMTP
1373 MAIL command, and the trusted user did not use -F to supply a sender
1374 name, $originator_name was incorrectly used when constructing a From:
1375 header. Furthermore, $originator_name was used for submission mode
1376 messages from external hosts without From: headers in a similar way,
1377 which is clearly wrong.
1379 PH/37 Added control=suppress_local_fixups.
1381 PH/38 When log_selector = +received_sender was set, and the addition of the
1382 sender made the log line's construction buffer exactly full, or one byte
1383 less than full, an overflow happened when the terminating "\n" was
1386 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1387 when the result of a list match is failure because a DNS lookup failed.
1389 PH/40 RM_COMMAND is now used in the building process.
1391 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1392 the "build-* directories that it finds.
1394 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1395 address in a domain literal was a prefix of an interface address.
1397 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1398 when verifying a sender address, unless rewrite_headers is false.
1400 PH/44 Wrote a long comment about why errors_to addresses are verified as
1401 recipients, not senders.
1403 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1404 the ratelimit ACL was added.
1406 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1408 PH/46 Added extra information about PostgreSQL errors to the error string.
1410 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1411 avoiding the need to install special zones in a real server. This is
1412 backwards compatible; if it can't find the fake resolver, it drops back.
1413 Thus, both old and new test suites can be run.
1415 TF/02 Added util/ratelimit.pl
1417 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1418 clock is set back in time.
1420 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1421 Candler <B.Candler@pobox.com>.
1423 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1424 OK for addresses that are the result of redirections.
1426 PH/48 A number of further additions for the benefit of the new test suite,
1427 including a fake gethostbyname() that interfaces to the fake DNS resolver
1430 TF/06 The fix for widen_domains has also been applied to qualify_single and
1431 search_parents which are the other dnslookup options that can cause
1434 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1437 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1439 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1440 service" if it encountered an MX record with an empty target hostname.
1441 The message is now "an MX or SRV record indicated no SMTP service".
1443 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1444 if the DNS of the sending site is misconfigured. This is quite a
1445 common situation. This change restores the behaviour of treating a
1446 helo verification defer as a failure.
1448 PH/51 If self=fail was set on a router, the bounce message did not include the
1449 actual error message.
1455 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1457 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1458 command itself ended in error, the underlying message about the transport
1459 filter timeout was being overwritten with the pipe command error. Now the
1460 underlying error message should be appended to the second error message.
1462 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1464 PH/02 Reduce the amount of output that "make" produces by default. Full output
1465 can still be requested.
1467 PH/03 The warning log line about a condition test deferring for a "warn" verb
1468 was being output only once per connection, rather than after each
1469 occurrence (because it was using the same function as for successful
1470 "warn" verbs). This seems wrong, so I have changed it.
1472 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1473 it should not have, which might have caused a crash in the right
1474 circumstances, but probably never did.
1476 PH/04 Installed a modified version of Tony Finch's patch to make submission
1477 mode fix the return path as well as the Sender: header line, and to
1478 add a /name= option so that you can make the user's friendly name appear
1481 TF/03 Added the control = fakedefer ACL modifier.
1483 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1484 Mark Lowes for thorough testing.
1486 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1488 TK/03 Merged latest SRS patch from Miles Wilton.
1490 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1491 with the definition in sysexits.h (which is #included earlier).
1492 Fortunately, Exim does not actually use EX_OK. The code used to try to
1493 preserve the sysexits.h value, by assumimg that macro definitions were
1494 scanned for macro replacements. I have been disabused of this notion,
1495 so now the code just undefines EX_OK before #including unistd.h.
1497 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1498 in the smtp transport. When a block could not be written in a single
1499 write() function, the timeout was being re-applied to each part-write.
1500 This seems wrong - if the receiver was accepting one byte at a time it
1501 would take for ever. The timeout is now adjusted when this happens. It
1502 doesn't have to be particularly precise.
1504 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1505 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1507 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1508 without a domain if the check to <postmaster@domain> fails.
1510 SC/01 Eximstats: added -xls and the ability to specify output files
1511 (patch written by Frank Heydlauf).
1513 SC/02 Eximstats: use FileHandles for outputing results.
1515 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1517 SC/04 Eximstats: fixed display of large numbers with -nvr option
1519 SC/05 Eximstats: fixed merging of reports with empty tables.
1521 SC/06 Eximstats: added the -include_original_destination flag
1523 SC/07 Eximstats: removed tabs and trailing whitespace.
1525 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1527 TK/06 MBOX spool code: Add real "From " MBOX separator line
1528 so the .eml file is really in mbox format (even though
1529 most programs do not really care). Patch from Alex Miller.
1531 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1532 The latter is generated from $received_to and is only set if the
1533 message has one envelope recipient. SA can use these headers,
1534 obviously out-of-the-box. Patch from Alex Miller.
1536 PH/08 The ${def test on a variable was returning false if the variable's
1537 value was "0", contrary to what the specification has always said!
1538 The result should be true unless the variable is empty.
1540 PH/09 The syntax error of a character other than { following "${if
1541 def:variable_name" (after optional whitespace) was not being diagnosed.
1542 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1543 accidental colon was present, for example, could give incorrect results.
1545 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1546 result is used (not including appendfile, where other changes are about
1549 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1550 This involved changing a lot of size variables from int to off_t. It
1551 should work with maildirs and everything.
1553 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1554 spamd dying while we are connected to it.
1556 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1559 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1560 function for BDB 4.3.
1562 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1564 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1565 the sources for PCRE have been re-arranged and more files are now
1568 PH/15 The code I had for printing potentially long long variables in PH/11
1569 above was not the best (it lost precision). The length of off_t variables
1570 is now inspected at build time, and an appropriate printing format (%ld
1571 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1572 to be "long long int" or "long int". This is needed for the internal
1573 formatting function string_vformat().
1575 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1576 the configuration file to be ":syslog", then the script "guesses" where
1577 the logs files are, rather than using the compiled in default. In our
1578 case the guess is not the same as the compiled default, so the script
1579 suddenly stopped working when I started to use syslog. The patch checks
1580 to see if log_file_path is "". If so, it attempts to read it from exim
1581 with no configuration file to get the compiled in version, before it
1582 falls back to the previous guessing code."
1584 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1585 implementing BATV in an Exim configuration. See NewStuff for the gory
1588 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1589 Makefile that are specific to HP-UX.
1591 PH/18 If the "use_postmaster" option was set for a recipient callout together
1592 with the "random" option, the postmaster address was used as the MAIL
1593 FROM address for the random test, but not for the subsequent recipient
1594 test. It is now used for both.
1596 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1597 patch removes a few documentation additions to RFC 3028, because the
1598 latest draft now contains them. It adds the new en;ascii-case comparator
1599 and a new error check for 8bit text in MIME parts. Comparator and
1600 require names are now matched exactly. I enabled the subaddress
1601 extension, but it is not well tested yet (read: it works for me)."
1603 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1604 rework some of the code of TK/09 above to avoid the hardwired use of
1605 "%lld" and "long long". Replaced the call to snprintf() with a call to
1608 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1609 records point to non-existent hosts", "retry timeout exceeded", and
1610 "retry time not reached for any host after a long failure period".
1612 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1613 experimental DomainKeys support:
1615 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1616 (2) On an error such as an illegally used "control", the wrong name for
1617 the control was given.
1619 These problems did NOT occur unless DomainKeys support was compiled.
1621 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1623 PH/24 Added ${if match_ip condition.
1625 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1626 databases so that it will be absolutely obvious if a crash occurs in the
1627 DB library. This is a regular occurrence (often caused by mis-matched
1630 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1631 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1632 picked up on a user's system that detects such things. There doesn't seem
1633 to be a gcc warning option for this - only an attribute that has to be
1634 put on the function's prototype. It seems that in Fedora Core 4 they have
1635 set this on a number of new functions. No doubt there will be more in due
1638 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1639 specify a transport. However, if an address that was verified by such a
1640 router was the subject of a callout, Exim crashed because it tried to
1641 read the rcpt_include_affixes from the non-existent transport. Now it
1642 just assumes that the setting of that option is false. This bug was
1643 introduced by 4.51/PH/31.
1645 PH/28 Changed -d+all to exclude +memory, because that information is very
1646 rarely of interest, but it makes the output a lot bigger. People tend to
1647 do -d+all out of habit.
1649 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1650 code in os-type was giving problems when libc.so lives in lib64, like on
1653 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1654 aren't the modern standard, and it seems that some systems' include files
1655 don't always have them. Exim was already checking for some of the newer
1656 ones like T_AAAA, and defining it itself. I've added checks for all the
1657 record types that Exim uses.
1659 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1660 not automatically generating a new one, as it is supposed to. This
1661 prevented TLS from working. If the file did exist, but contained invalid
1662 data, a new version was generated, as expected. It was only the case of a
1663 non-existent file that was broken.
1665 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1666 with a change in libdomainkeys > 0.64.
1668 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1669 from DNS. If the selector record carries the flag, it now has
1670 precedence over the domain-wide flag.
1672 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1674 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1675 the use of a port name that isn't defined in /etc/services) occurred, the
1676 message was deferred as in a normal delivery, and thus remained on the
1677 spool, instead of being failed because of the mua_wrapper setting. This
1678 is now fixed, and I tidied up some of the mua_wrapper messages at the
1681 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1682 the messages in a hash of arrays rather than using individual hashes.
1683 This is a bit cleaner and results in dramatic memory savings, albeit
1684 at a slight CPU cost.
1686 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1687 as requested by Marc Sherman.
1689 SC/10 Eximstats: added histograms for user specified patterns as requested
1692 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1694 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1695 fopen() in the content-scanning modules that did not already have it.
1701 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1702 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1704 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1706 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1708 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1709 to test Sieve filters that use "vacation".
1711 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1712 that changes the way the GnuTLS parameters are stored in the cache file.
1713 The new format can be generated externally. For backward compatibility,
1714 if the data in the cache doesn't make sense, Exim assumes it has read an
1715 old-format file, and it generates new data and writes a new file. This
1716 means that you can't go back to an older release without removing the
1719 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1720 work if there are any delivery delays because "one_time" forces the
1721 parent to be marked "delivered", so its unseen clone is never tried
1722 again. For this reason, Exim now forbids the simultaneous setting of
1725 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1726 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1727 entry if you want to know the details. The fix, however, neglected to
1728 consider the case when local delivery batching is involved. The test for
1729 "previously delivered" was not happening when checking to see if an
1730 address could be batched with a previous (undelivered) one; under
1731 certain circumstances this could lead to multiple deliveries to the same
1734 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1735 in its include files, and this causes problems building Exim.
1737 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1738 header_syntax) but Exim was just ignoring anything given after a slash.
1739 In particular, this caused confusion with an attempt to use "verify =
1740 reverse_host_lookup/defer_ok". An error is now given when options are
1741 supplied for verify items that do not have them. (Maybe reverse_host_
1742 lookup should have a defer_ok option, but that's a different point.)
1744 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1745 defined by RFC 821) to 2048, because there were problems with some AUTH
1746 commands, and RFC 1869 says the size should be increased for extended
1747 SMTP commands that take arguments.
1749 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1752 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1753 "unknown" error; now it says that the functionality isn't in the binary.
1755 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1756 an address' error message when a string expansion fails (syntax or
1757 whatever). Otherwise the password may appear in the log. Following change
1758 PH/42 below, there is no longer a chance of it appearing in a bounce
1761 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1763 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1764 of its addresses was ever tried. (Bugzilla bug #2.)
1766 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1767 the result incorrectly in the debug output. (It correctly added a newline
1768 to what was transported.)
1770 TF/01 Added $received_time.
1772 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1773 commented out examples of how to interface to a virus scanner and to
1774 SpamAssassin. Also added commented examples of av_scanner and
1775 spamd_address settings.
1777 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1778 and controls are allowed in which ACLs. There were a couple of minor
1779 errors. Some of the entries in the conditions table (which is a table of
1780 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1781 negation of where the condition IS allowed.
1783 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1785 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1786 header file does not have a version number, so I've had to invent a new
1787 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1788 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1789 radiusclient), but it was contributed by a Radius user.
1791 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1792 files or directories, for OpenSSL.
1794 PH/19 When an Exim process that is running as root has to create an Exim log
1795 file, it does so in a subprocess that runs as exim:exim so as to get the
1796 ownership right at creation (otherwise, other Exim processes might see
1797 the file with the wrong ownership). There was no test for failure of this
1798 fork() call, which would lead to the process getting stuck as it waited
1799 for a non-existent subprocess. Forks do occasionally fail when resources
1800 run out. I reviewed all the other calls to fork(); they all seem to check
1803 PH/20 When checking for unexpected SMTP input at connect time (before writing
1804 the banner), Exim was not dealing correctly with a non-positive return
1805 from the read() function. If the client had disconnected by this time,
1806 the result was a log entry for a synchronization error with an empty
1807 string after "input=" when read() returned zero. If read() returned -1
1808 (an event I could not check), uninitialized data bytes were printed.
1809 There were reports of junk text (parts of files, etc) appearing after
1812 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1814 PH/22 Added support for macro redefinition, and (re)definition in between
1815 driver and ACL definitions.
1817 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1818 forgetting to use the resulting value; it was using the unexpanded value.
1820 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1821 hadn't been configured. The fix is from Juergen Kreileder, who
1822 understands it better than I do:
1824 "Here's what I see happening with three configured cyrus_sasl
1825 authenticators configured (plain, login, cram-md5):
1827 On startup auth_cyrus_sasl_init() gets called for each of these.
1828 This means three calls to sasl_listmech() without a specified mech_list.
1829 => SASL tests which mechs of all available mechs actually work
1830 => three warnings about OTP not working
1831 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1833 With the patch, sasl_listmech() also gets called three times. But now
1834 SASL's mech_list option is set to the server_mech specified in the the
1835 authenticator. Or in other words, the answer from sasl_listmech()
1836 gets limited to just the mech you're testing for (which is different
1838 => the return list contains just 'plain' or 'login', 'cram-md5' or
1839 nothing depending on the value of ob->server_mech.
1841 I've just tested the patch: Authentication still works fine,
1842 unavailable mechs specified in the exim configuration are still
1843 caught, and the auth.log warnings about OTP are gone."
1845 PH/25 When debugging is enabled, the contents of the command line are added
1846 to the debugging output, even when log_selector=+arguments is not
1849 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1850 answer is "GNU", and only if the return is "GNU/something" is the answer
1853 PH/27 $acl_verify_message is now set immediately after the failure of a
1854 verification in an ACL, and so is available in subsequent modifiers. In
1855 particular, the message can be preserved by coding like this:
1857 warn !verify = sender
1858 set acl_m0 = $acl_verify_message
1860 Previously, $acl_verify_message was set only while expanding "message"
1861 and "log_message" when a very denied access.
1863 PH/28 Modified OS/os.c-Linux with
1865 -#ifndef OS_LOAD_AVERAGE
1866 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1868 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1869 nomenclature these days.)
1871 PH/29 Installed patch from the Sieve maintainer that adds the options
1872 sieve_useraddress and sieve_subaddress to the redirect router.
1874 PH/30 In these circumstances:
1875 . Two addresses routed to the same list of hosts;
1876 . First host does not offer TLS;
1877 . First host accepts first address;
1878 . First host gives temporary error to second address;
1879 . Second host offers TLS and a TLS session is established;
1880 . Second host accepts second address.
1881 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1882 and peerdn, if requested) that were in fact used only for the second
1885 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1886 attention to any local part prefix or suffix that was matched by the
1887 router that accepted the address. It now behaves in the same way as it
1888 does for delivery: the affixes are removed from the local part unless
1889 rcpt_include_affixes is set on the transport.
1891 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1892 timeout during the DATA phase of an incoming message.
1894 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1895 applied a patch sanctioned by the Sieve maintainer.
1897 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1898 the uid or gid is negative. A case of a negative gid caused this to be
1899 noticed. The fix allows for either to be negative.
1901 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1902 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1905 PH/36 The change PH/12 above was broken. Fixed it.
1907 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1908 the grounds that routing the same address twice would always produce the
1909 same answer. This might have been true once, but it is certainly no
1910 longer true now. Routing a child address may depend on the previous
1911 routing that produced that child. Some complicated redirection strategies
1912 went wrong when messages had multiple recipients, and made Exim's
1913 behaviour dependent on the order in which the addresses were given.
1915 I have moved the duplicate checking until after the routing is complete.
1916 Exim scans the addresses that are assigned to local and remote
1917 transports, and removes any duplicates. This means that more work will be
1918 done, as duplicates will always all be routed, but duplicates are
1919 presumably rare, so I don't expect this is of any significance.
1921 For deliveries to pipes, files, and autoreplies, the duplicate checking
1922 still happens during the routing process, since they are not going to be
1925 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1926 It corrects a timeout issue with spamd. This is Ian's comment: "The
1927 background is that sometimes spamd either never reads data from a
1928 connection it has accepted, or it never writes response data. The exiscan
1929 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1930 blindly assumes that writes won't block so it may never time out."
1932 PH/39 Allow G after quota size as well as K and M.
1934 PH/40 The value set for $authenticated_id in an authenticator may not contain
1935 binary zeroes or newlines because the value is written to log lines and
1936 to spool files. There was no check on this. Now the value is run through
1937 the string_printing() function so that such characters are converted to
1938 printable escape sequences.
1940 PH/41 $message_linecount is a new variable that contains the total number of
1941 lines in the message. Compare $body_linecount, which is the count for the
1944 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1945 bounce and delay warning messages, except in certain special cases, which
1948 (a) An SMTP error message from a remote host;
1949 (b) A message specified in a :fail: redirection;
1950 (c) A message specified in a "fail" command in a system filter;
1951 (d) A message specified in a FAIL return from the queryprogram router;
1952 (e) A message specified by the cannot_route_message router option.
1954 In these cases only, Exim does include the error details in bounce and
1955 warning messages. There are also a few cases where bland messages such
1956 as "unrouteable address" or "local delivery error" are given.
1958 PH/43 $value is now also set for the "else" part of a ${run expansion.
1960 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1961 being worked on, but at least Exim now implements the latest version to
1964 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1965 process to complete was treated as a delivery failure, a timeout while
1966 writing the message to the pipe was logged, but erroneously treated as a
1967 successful delivery. Such timeouts include transport filter timeouts. For
1968 consistency with the overall process timeout, these timeouts are now
1969 treated as errors, giving rise to delivery failures by default. However,
1970 there is now a new Boolean option for the pipe transport called
1971 timeout_defer, which, if set TRUE, converts the failures into defers for
1972 both kinds of timeout. A transport filter timeout is now identified in
1975 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1976 systems where "make" and "gmake" are different, calling "gmake" at top
1977 level broke things. I've arranged for the value of $(MAKE) to be passed
1978 from the Makefile to this script so that it can call the same version of
1982 A note about Exim versions 4.44 and 4.50
1983 ----------------------------------------
1985 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1986 changes of various kinds. As a consequence, a big documentation update was
1987 needed. This delayed the release for rather longer than seemed good, especially
1988 in the light of a couple of (minor) security issues. Therefore, the changes
1989 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1990 release. So 4.44 and 4.50 are in effect two different branches that both start
1993 I have left the 4.50 change log unchanged; it contains all the changes since
1994 4.43. The change log for 4.44 is below; many of its items are identical to
1995 those for 4.50. This seems to be the most sensible way to preserve the
1996 historical information.
2002 1. Minor wording change to the doc/README.SIEVE file.
2004 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2005 computation of the current number of files was incorrect.
2007 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2008 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2011 4. Give more explanation in the error message when the command for a transport
2012 filter fails to execute.
2014 5. There are several places where Exim runs a non-Exim command in a
2015 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2016 was being done only for the command run by the queryprogram router. It is
2017 now done for all such subprocesses. The other cases are: ${run, transport
2018 filters, and the commands run by the lmtp and pipe transports.
2020 6. Added CONFIGURE_GROUP build-time option.
2022 7. Some older OS have a limit of 256 on the maximum number of file
2023 descriptors. Exim was using setrlimit() to set 1000 as a large value
2024 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2025 systems. I've change it so that if it can't get 1000, it tries for 256.
2027 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2028 was an oversight, and furthermore, ever since the addition of extra
2029 controls (e.g. 4.43/32), the checks on when to allow different forms of
2030 "control" were broken. There should now be diagnostics for all cases when a
2031 control that does not make sense is encountered.
2033 9. Added the /retain_sender option to "control=submission".
2035 10. $recipients is now available in the predata ACL (oversight).
2037 11. Tidy the search cache before the fork to do a delivery from a message
2038 received from the command line. Otherwise the child will trigger a lookup
2039 failure and thereby defer the delivery if it tries to use (for example) a
2040 cached ldap connection that the parent has called unbind on.
2042 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2043 of $address_data from the recipient verification was clobbered by the
2044 sender verification.
2046 13. The value of address_data from a sender verification is now available in
2047 $sender_address_data in subsequent conditions in the ACL statement.
2049 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2051 15. Added a new option "connect=<time>" to callout options, to set a different
2054 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2055 was its contents. (It was OK if the option was not defined at all.)
2057 17. A "Completed" log line is now written for messages that are removed from
2058 the spool by the -Mrm option.
2060 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2061 information about exactly what failed.
2063 19. Added -dd to debug only the daemon process.
2065 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2066 handles timeouts, both on the server side and network timeouts. Renamed the
2067 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2069 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2070 It is now set to "smtps".
2072 22. $host_address is now set to the target address during the checking of
2073 ignore_target_hosts.
2075 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2076 being passed; this would have caused $sender_host_name to have been used if
2077 matching the list had actually called for a host name (not very likely,
2078 since this list is usually IP addresses). A host name is now passed as
2081 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2082 code that specifies a non-restarting handler (typically sigaction() in
2083 modern systems) in an attempt to fix a rare and obscure crash bug.
2085 25. Narrowed the window for a race in the daemon that could cause it to ignore
2086 SIGCHLD signals. This is not a major problem, because they are used only to
2087 wake it up if nothing else does.
2089 26. A malformed maildirsize file could cause Exim to calculate negative values
2090 for the mailbox size or file count. Odd effects could occur as a result.
2091 The maildirsize information is now recalculated if the size or filecount
2094 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2095 support for a long time. Removed HAVE_SYS_VFS_H.
2097 28. Installed the latest version of exipick from John Jetmore.
2099 29. In an address list, if the pattern was not a regular expression, an empty
2100 subject address (from a bounce message) matched only if the pattern was an
2101 empty string. Non-empty patterns were not even tested. This was the wrong
2102 because it is perfectly reasonable to use an empty address as part of a
2103 database query. An empty address is now tested by patterns that are
2104 lookups. However, all the other forms of pattern expect the subject to
2105 contain a local part and a domain, and therefore, for them, an empty
2106 address still always fails if the pattern is not itself empty.
2108 30. Exim went into a mad DNS loop when attempting to do a callout where the
2109 host was specified on an smtp transport, and looking it up yielded more
2110 than one IP address.
2112 31. Re-factored the code for checking spool and log partition space into a
2113 function that finds that data and another that does the check. The former
2114 is then used to implement four new variables: $spool_space, $log_space,
2115 $spool_inodes, and $log_inodes.
2117 32. The RFC2047 encoding function was originally intended for short strings
2118 such as real names; it was not keeping to the 75-character limit for
2119 encoded words that the RFC imposes. It now respects the limit, and
2120 generates multiple encoded words if necessary. To be on the safe side, I
2121 have increased the buffer size for the ${rfc2047: expansion operator from
2124 33. It is now permitted to omit both strings after an "if" condition; if the
2125 condition is true, the result is "true". As before, when the second string
2126 is omitted, a false condition yields an empty string. This makes it less
2127 cumbersome to write custom ACL and router conditions.
2129 34. Failure to deliver a bounce message always caused it to be frozen, even if
2130 there was an errors_to setting on the router. The errors_to setting is now
2133 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2134 canonical form (fully expanded) before being placed in
2135 $sender_host_address.
2137 36. The table in the code that translates DNS record types into text (T_A to
2138 "A" for instance) was missing entries for NS and CNAME. It is just possible
2139 that this could have caused confusion if both these types were looked up
2140 for the same domain, because the text type is used as part of Exim's
2141 per-process caching. But the chance of anyone hitting this buglet seems
2144 37. The dnsdb lookup has been extended in a number of ways.
2146 (1) There is a new type, "zns", which walks up the domain tree until it
2147 finds some nameserver records. It should be used with care.
2149 (2) There is a new type, "mxh", which is like "mx" except that it returns
2150 just the host names, not the priorities.
2152 (3) It is now possible to give a list of domains (or IP addresses) to be
2153 looked up. The behaviour when one of the lookups defers can be
2154 controlled by a keyword.
2156 (4) It is now possible to specify the separator character for use when
2157 multiple records are returned.
2159 38. The dnslists ACL condition has been extended: it is now possible to supply
2160 a list of IP addresses and/or domains to be looked up in a particular DNS
2163 39. Added log_selector=+queue_time_overall.
2165 40. When running the queue in the test harness, wait just a tad after forking a
2166 delivery process, to get repeatability of debugging output.
2168 41. Include certificate and key file names in error message when GnuTLS fails
2169 to set them up, because the GnuTLS error message doesn't include the name
2170 of the failing file when there is a problem reading it.
2172 42. Allow both -bf and -bF in the same test run.
2174 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2176 44. The "Exiscan patch" is now merged into the mainline Exim source.
2178 45. Sometimes the final signoff response after QUIT could fail to get
2179 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2180 < 0 before doing a fflush(). This bug looks as though it goes back to the
2181 introduction of TLS in release 3.20, but "sometimes" must have been rare
2182 because the tests only now provoked it.
2184 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2185 (this can affect the format of dates).
2187 47. exim_tidydb, when checking for the continued existence of a message for
2188 which it has found a message-specific retry record, was not finding
2189 messages that were in split spool directories. Consequently, it was
2190 deleting retry records that should have stayed in existence.
2192 48. Steve fixed some bugs in eximstats.
2194 49. The SPA authentication driver was not abandoning authentication and moving
2195 on to the next authenticator when an expansion was forced to fail,
2196 contradicting the general specification for all authenticators. Instead it
2197 was generating a temporary error. It now behaves as specified.
2199 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2200 (the order specifies the preference for clients). The order is now AES256,
2201 AES128, 3DES, ARCFOUR128.
2203 51. Small patch to Sieve code - explicitly set From: when generating an
2206 52. Exim crashed if a remote delivery caused a very long error message to be
2207 recorded - for instance if somebody sent an entire SpamAssassin report back
2208 as a large number of 550 error lines. This bug was coincidentally fixed by
2209 increasing the size of one of Exim's internal buffers (big_buffer) that
2210 happened as part of the Exiscan merge. However, to be on the safe side, I
2211 have made the code more robust (and fixed the comments that describe what
2214 53. Now that there can be additional text after "Completed" in log lines (if
2215 the queue_time_overall log selector is set), a one-byte patch to exigrep
2216 was needed to allow it to recognize "Completed" as not the last thing in
2219 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2220 patch that reportedly fixes this has been added. I am not expert enough to
2221 create a test for it. This is what the patch creator wrote:
2223 "I found a little strange behaviour of ldap code when working with
2224 Windows 2003 AD Domain, where users was placed in more than one
2225 Organization Units. When I tried to give exim partial DN, the exit code
2226 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2227 But simultaneously result of request was absolutely normal ldap result,
2228 so I produce this patch..."
2230 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2231 REFERENCE, so I have modified the code to exclude the patch when that macro
2234 55. Some experimental protocols are using DNS PTR records for new purposes. The
2235 keys for these records are domain names, not reversed IP addresses. The
2236 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2237 leaves it alone. Component reversal etc. now happens only for IP addresses.
2240 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2242 57. Double the size of the debug message buffer (to 2048) so that more of very
2243 long debug lines gets shown.
2245 58. The exicyclog utility now does better if the number of log files to keep
2246 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2248 59. Two changes related to the smtp_active_hostname option:
2250 (1) $smtp_active_hostname is now available as a variable.
2251 (2) The default for smtp_banner uses $smtp_active_hostname instead
2252 of $primary_hostname.
2254 60. The host_aton() function is supposed to be passed a string that is known
2255 to be a valid IP address. However, in the case of IPv6 addresses, it was
2256 not checking this. This is a hostage to fortune. Exim now panics and dies
2257 if the condition is not met. A case was found where this could be provoked
2258 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2259 components; fortuitously, this particular loophole had already been fixed
2260 by change 4.50/55 above.
2262 If there are any other similar loopholes, the new check in host_aton()
2263 itself should stop them being exploited. The report I received stated that
2264 data on the command line could provoke the exploit when Exim was running as
2265 exim, but did not say which command line option was involved. All I could
2266 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2267 running as the user.
2270 61. There was a buffer overflow vulnerability in the SPA authentication code
2271 (which came originally from the Samba project). I have added a test to the
2272 spa_base64_to_bits() function which I hope fixes it.
2275 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2276 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2278 63. The daemon start-up calls getloadavg() while still root for those OS that
2279 need the first call to be done as root, but it missed one case: when
2280 deliver_queue_load_max is set with deliver_drop_privilege. This is
2281 necessary for the benefit of the queue runner, because there is no re-exec
2282 when deliver_drop_privilege is set.
2284 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2285 This has been fixed.
2287 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2288 was in use, was not putting the data itself into the right store pool;
2289 consequently, it could be overwritten for a subsequent message in the same
2290 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2293 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2295 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2296 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2297 the function were treating the return as a boolean value, which happened to
2298 work because 0=false and not-0=true, but is not correct code.
2300 68. The host_aton() function was not handling scoped IPv6 addresses (those
2301 with, for example, "%eth0" on the end) correctly.
2303 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2304 negated items (that is, ~something) in unsigned ints. Some compilers
2305 apparently mutter when there is no cast.
2307 70. If an address verification called from an ACL failed, and did not produce a
2308 user-specific message (i.e. there was only a "system" message), nothing was
2309 put in $acl_verify_message. In this situation, it now puts the system
2312 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2313 session; change 4.31/43 added the unwanted input to the log line - except
2314 that it did not do this in the start of session case. It now does.
2316 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2317 This is wrong and can cause the other end to generate a synchronization
2318 error if it is another Exim or anything else that does the synchronization
2319 check. A QUIT command is no longer sent after a timeout.
2321 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2322 during host lookups.
2324 74. The defer_ok option of callout verification was not working if it was used
2325 when verifying addresses in header lines, that is, for this case:
2327 verify = header_sender/callout=defer_ok
2329 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2330 those file descriptors could be used for SMTP connections. If anything
2331 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2332 could be sent to the SMTP client, causing chaos. The daemon now opens
2333 stdin, stdout, and stderr to /dev/null when it puts itself into the
2336 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2337 log by default. The user can override this with suitable Perl magic.
2339 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2340 the log message when discard triggers, was not working for the DATA ACL or
2341 for the non-SMTP ACL.
2343 78. Error message wording change in sieve.c.
2345 79. If smtp_accept_max_per_host was set, the number of connections could be
2346 restricted to fewer than expected, because the daemon was trying to set up
2347 a new connection before checking whether the processes handling previous
2348 connections had finished. The check for completed processes is now done
2349 earlier. On busy systems, this bug wouldn't be noticed because something
2350 else would have woken the daemon, and it would have reaped the completed
2353 80. If a message was submitted locally by a user whose login name contained one
2354 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2355 It caused a spool format error. I have fixed the spool reading code. A
2356 related problem was that the "from" clause in the Received: line became
2357 illegal because of the space(s). It is now covered by ${quote_local_part.
2359 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2362 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2364 83. Patch from Sieve maintainer to fix unterminated string problem in
2365 "vacation" handling.
2367 84. Some minor changes to the Linux configuration files to help with other
2368 OS variants using glibc.
2370 85. One more patch for Sieve to update vacation handling to latest spec.
2373 ----------------------------------------------------
2374 See the note above about the 4.44 and 4.50 releases.
2375 ----------------------------------------------------
2381 1. Change 4.43/35 introduced a bug that caused file counts to be
2382 incorrectly computed when quota_filecount was set in an appendfile
2385 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2386 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2389 3. Give more explanation in the error message when the command for a transport
2390 filter fails to execute.
2392 4. There are several places where Exim runs a non-Exim command in a
2393 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2394 was being done only for the command run by the queryprogram router. It is
2395 now done for all such subprocesses. The other cases are: ${run, transport
2396 filters, and the commands run by the lmtp and pipe transports.
2398 5. Some older OS have a limit of 256 on the maximum number of file
2399 descriptors. Exim was using setrlimit() to set 1000 as a large value
2400 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2401 systems. I've change it so that if it can't get 1000, it tries for 256.
2403 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2404 was an oversight, and furthermore, ever since the addition of extra
2405 controls (e.g. 4.43/32), the checks on when to allow different forms of
2406 "control" were broken. There should now be diagnostics for all cases when a
2407 control that does not make sense is encountered.
2409 7. $recipients is now available in the predata ACL (oversight).
2411 8. Tidy the search cache before the fork to do a delivery from a message
2412 received from the command line. Otherwise the child will trigger a lookup
2413 failure and thereby defer the delivery if it tries to use (for example) a
2414 cached ldap connection that the parent has called unbind on.
2416 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2417 of $address_data from the recipient verification was clobbered by the
2418 sender verification.
2420 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2421 was its contents. (It was OK if the option was not defined at all.)
2423 11. A "Completed" log line is now written for messages that are removed from
2424 the spool by the -Mrm option.
2426 12. $host_address is now set to the target address during the checking of
2427 ignore_target_hosts.
2429 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2430 being passed; this would have caused $sender_host_name to have been used if
2431 matching the list had actually called for a host name (not very likely,
2432 since this list is usually IP addresses). A host name is now passed as
2435 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2436 code that specifies a non-restarting handler (typically sigaction() in
2437 modern systems) in an attempt to fix a rare and obscure crash bug.
2439 15. Narrowed the window for a race in the daemon that could cause it to ignore
2440 SIGCHLD signals. This is not a major problem, because they are used only to
2441 wake it up if nothing else does.
2443 16. A malformed maildirsize file could cause Exim to calculate negative values
2444 for the mailbox size or file count. Odd effects could occur as a result.
2445 The maildirsize information is now recalculated if the size or filecount
2448 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2449 support for a long time. Removed HAVE_SYS_VFS_H.
2451 18. Updated exipick to current release from John Jetmore.
2453 19. Allow an empty sender to be matched against a lookup in an address list.
2454 Previously the only cases considered were a regular expression, or an
2457 20. Exim went into a mad DNS lookup loop when doing a callout where the
2458 host was specified on the transport, if the DNS lookup yielded more than
2461 21. The RFC2047 encoding function was originally intended for short strings
2462 such as real names; it was not keeping to the 75-character limit for
2463 encoded words that the RFC imposes. It now respects the limit, and
2464 generates multiple encoded words if necessary. To be on the safe side, I
2465 have increased the buffer size for the ${rfc2047: expansion operator from
2468 22. Failure to deliver a bounce message always caused it to be frozen, even if
2469 there was an errors_to setting on the router. The errors_to setting is now
2472 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2473 canonical form (fully expanded) before being placed in
2474 $sender_host_address.
2476 24. Updated eximstats to version 1.33
2478 25. Include certificate and key file names in error message when GnuTLS fails
2479 to set them up, because the GnuTLS error message doesn't include the name
2480 of the failing file when there is a problem reading it.
2482 26. Expand error message when OpenSSL has problems setting up cert/key files.
2485 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2486 (this can affect the format of dates).
2488 28. exim_tidydb, when checking for the continued existence of a message for
2489 which it has found a message-specific retry record, was not finding
2490 messages that were in split spool directories. Consequently, it was
2491 deleting retry records that should have stayed in existence.
2493 29. eximstats updated to version 1.35
2494 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2495 1.35 - bugfix such that pie charts by volume are generated correctly
2497 30. The SPA authentication driver was not abandoning authentication and moving
2498 on to the next authenticator when an expansion was forced to fail,
2499 contradicting the general specification for all authenticators. Instead it
2500 was generating a temporary error. It now behaves as specified.
2502 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2503 (the order specifies the preference for clients). The order is now AES256,
2504 AES128, 3DES, ARCFOUR128.
2506 31. Small patch to Sieve code - explicitly set From: when generating an
2509 32. Exim crashed if a remote delivery caused a very long error message to be
2510 recorded - for instance if somebody sent an entire SpamAssassin report back
2511 as a large number of 550 error lines. This bug was coincidentally fixed by
2512 increasing the size of one of Exim's internal buffers (big_buffer) that
2513 happened as part of the Exiscan merge. However, to be on the safe side, I
2514 have made the code more robust (and fixed the comments that describe what
2517 33. Some experimental protocols are using DNS PTR records for new purposes. The
2518 keys for these records are domain names, not reversed IP addresses. The
2519 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2520 leaves it alone. Component reversal etc. now happens only for IP addresses.
2523 34. The host_aton() function is supposed to be passed a string that is known
2524 to be a valid IP address. However, in the case of IPv6 addresses, it was
2525 not checking this. This is a hostage to fortune. Exim now panics and dies
2526 if the condition is not met. A case was found where this could be provoked
2527 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2528 components; fortuitously, this particular loophole had already been fixed
2529 by change 4.50/55 or 4.44/33 above.
2531 If there are any other similar loopholes, the new check in host_aton()
2532 itself should stop them being exploited. The report I received stated that
2533 data on the command line could provoke the exploit when Exim was running as
2534 exim, but did not say which command line option was involved. All I could
2535 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2536 running as the user.
2539 35. There was a buffer overflow vulnerability in the SPA authentication code
2540 (which came originally from the Samba project). I have added a test to the
2541 spa_base64_to_bits() function which I hope fixes it.
2544 36. The daemon start-up calls getloadavg() while still root for those OS that
2545 need the first call to be done as root, but it missed one case: when
2546 deliver_queue_load_max is set with deliver_drop_privilege. This is
2547 necessary for the benefit of the queue runner, because there is no re-exec
2548 when deliver_drop_privilege is set.
2550 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2551 was in use, was not putting the data itself into the right store pool;
2552 consequently, it could be overwritten for a subsequent message in the same
2553 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2556 38. Sometimes the final signoff response after QUIT could fail to get
2557 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2558 < 0 before doing a fflush(). This bug looks as though it goes back to the
2559 introduction of TLS in release 3.20, but "sometimes" must have been rare
2560 because the tests only now provoked it.
2566 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2567 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2568 Now it returns an integer. A number of calls were still expecting a T/F
2569 return. Fortuitously, in all cases, the tests worked in OK situations,
2570 which is the norm. However, things would have gone wrong on any write
2571 failures on the smtp file descriptor. This function is used when sending
2572 messages over SMTP and also when doing verify callouts.
2574 2. When Exim is called to do synchronous delivery of a locally submitted
2575 message (the -odf or -odi options), it no longer closes stderr before doing
2578 3. Implemented the mua_wrapper option.
2580 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2582 5. Implemented the functions header_remove(), header_testname(),
2583 header_add_at_position(), and receive_remove_recipient(), and exported them
2586 6. If an ACL "warn" statement specified the addition of headers, Exim already
2587 inserted X-ACL-Warn: at the start if there was no header name. However, it
2588 was not making this test for the second and subsequent header lines if
2589 there were newlines in the string. This meant that an invalid header could
2590 be inserted if Exim was badly configured.
2592 7. Allow an ACL "warn" statement to add header lines at the start or after all
2593 the Received: headers, as well as at the end.
2595 8. Added the rcpt_4xx retry error code.
2597 9. Added postmaster_mailfrom=xxx to callout verification option.
2599 10. Added mailfrom=xxxx to the callout verification option, for verify=
2602 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2603 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2605 12. Inserted some casts to stop certain compilers warning when using pointer
2606 differences as field lengths or precisions in printf-type calls (mostly
2607 affecting debugging statements).
2609 13. Added optional readline() support for -be (dynamically loaded).
2611 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2612 same clock tick as a message's arrival, so that its received time was the
2613 same as the "first fail" time on the retry record, and that message
2614 remained on the queue past the ultimate address timeout, every queue runner
2615 would try a delivery (because it was past the ultimate address timeout) but
2616 after another failure, the ultimate address timeout, which should have then
2617 bounced the address, did not kick in. This was a "< instead of <=" error;
2618 in most cases the first failure would have been in the next clock tick
2619 after the received time, and all would be well.
2621 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2622 being recognized when the domain list was tested by the match_domain
2623 condition in an expansion string.
2625 16. Added the ${str2b64: operator.
2627 17. Exim was always calling setrlimit() to set a large limit for the number of
2628 processes, without checking whether the existing limit was already
2629 adequate. (It did check for the limit on file descriptors.) Furthermore,
2630 errors from getrlimit() and setrlimit() were being ignored. Now they are
2631 logged to the main and panic logs, but Exim does carry on, to try to do its
2632 job under whatever limits there are.
2634 18. Imported PCRE 5.0.
2636 19. Trivial typo in log message " temporarily refused connection" (the leading
2639 20. If the log selector return_path_on_delivery was set and an address was
2640 redirected to /dev/null, the delivery process crashed because it assumed
2641 that a return path would always be set for a "successful" delivery. In this
2642 case, the whole delivery is bypassed as an optimization, and therefore no
2645 21. Internal re-arrangement: the function for sending a challenge and reading
2646 a response while authentication was assuming a zero-terminated challenge
2647 string. It's now changed to take a pointer and a length, to allow for
2648 binary data in such strings.
2650 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2652 23. Exim was not respecting finduser_retries when seeking the login of the
2653 uid under which it was called; it was always trying 10 times. (The default
2654 setting of finduser_retries is zero.) Also, it was sleeping after the final
2655 failure, which is pointless.
2657 24. Implemented tls_on_connect_ports.
2659 25. Implemented acl_smtp_predata.
2661 26. If the domain in control=submission is set empty, Exim assumes that the
2662 authenticated id is a complete email address when it generates From: or
2663 Sender: header lines.
2665 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2666 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2667 chown and chgrp in /bin and hostname in /usr/bin.
2669 28. Exim was keeping the "process log" file open after each use, just as it
2670 does for the main log. This opens the possibility of it remaining open for
2671 long periods when the USR1 signal hits a daemon. Occasional processlog
2672 errors were reported, that could have been caused by this. Anyway, it seems
2673 much more sensible not to leave this file open at all, so that is what now
2676 29. The long-running daemon process does not normally write to the log once it
2677 has entered its main loop, and it closes the log before doing so. This is
2678 so that log files can straightforwardly be renamed and moved. However,
2679 there are a couple of unusual error situations where the daemon does write
2680 log entries, and I had neglected to close the log afterwards.
2682 30. The text of an SMTP error response that was received during a remote
2683 delivery was being truncated at 512 bytes. This is too short for some of
2684 the long messages that one sometimes sees. I've increased the limit to
2687 31. It is now possible to make retry rules that apply only when a message has a
2688 specific sender, in particular, an empty sender.
2690 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2691 it possible to be selective about when SMTP synchronization is enforced.
2693 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2695 32. Implemented hosts_connection_nolog.
2697 33. Added an ACL for QUIT.
2699 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2702 35. Added mailbox_size and mailbox_filecount to appendfile.
2704 36. Added control = no_multiline_responses to ACLs.
2706 37. There was a bug in the logic of the code that waits for the clock to tick
2707 in the case where the clock went backwards by a substantial amount such
2708 that the microsecond fraction of "now" was more than the microsecond
2709 fraction of "then" (but the whole seconds number was less).
2711 38. Added support for the libradius Radius client library this is found on
2712 FreeBSD (previously only the radiusclient library was supported).
2718 1. When certain lookups returned multiple values in the form name=value, the
2719 quoting of the values was not always being done properly. Specifically:
2720 (a) If the value started with a double quote, but contained no whitespace,
2722 (b) If the value contained whitespace other than a space character (i.e.
2723 tabs or newlines or carriage returns) it was not quoted.
2724 This fix has been applied to the mysql and pgsql lookups by writing a
2725 separate quoting function and calling it from the lookup code. The fix
2726 should probably also be applied to nisplus, ibase and oracle lookups, but
2727 since I cannot test any of those, I have not disturbed their existing code.
2729 2. A hit in the callout cache for a specific address caused a log line with no
2730 reason for rejecting RCPT. Now it says "Previous (cached) callout
2731 verification failure".
2733 3. There was an off-by-one bug in the queryprogram router. An over-long
2734 return line was truncated at 256 instead of 255 characters, thereby
2735 overflowing its buffer with the terminating zero. As well as fixing this, I
2736 have increased the buffer size to 1024 (and made a note to document this).
2738 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2739 when Exim is waiting for an SMTP response from a remote server, Exim
2740 restarts its select() call on the socket, thereby resetting its timeout.
2741 This is not a problem when such interrupts are rare. Somebody set up a cron
2742 job to run exiwhat every 2 minutes, which is less than the normal select()
2743 timeout (5 or 10 minutes). This meant that the select() timeout never
2744 kicked in because it was always reset. I have fixed this by comparing the
2745 time when an interrupt arrives with the time at the start of the first call
2746 to select(). If more time than the timeout has elapsed, the interrupt is
2747 treated as a timeout.
2749 5. Some internal re-factoring in preparation for the addition of Sieve
2750 extensions (by MH). In particular, the "personal" test is moved to a
2751 separate function, and given an option for scanning Cc: and Bcc: (which is
2752 not set for Exim filters).
2754 6. When Exim created an email address using the login of the caller as the
2755 local part (e.g. when creating a From: or Sender: header line), it was not
2756 quoting the local part when it contained special characters such as @.
2758 7. Installed new OpenBSD configuration files.
2760 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2761 try to make them clearer.
2763 9. Callout options, other than the timeout value, were being ignored when
2764 verifying sender addresses in header lines. For example, when using
2766 verify = header_sender/callout=no_cache
2768 the cache was (incorrectly) being used.
2770 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2771 only the Cygwin environment.
2773 11. When return_path_on_delivery was set as a log selector, if different remote
2774 addresses in the same message used different return paths and parallel
2775 remote delivery occurred, the wrong values would sometimes be logged.
2776 (Whenever a remote delivery process finished, the return path value from
2777 the most recently started remote delivery process was logged.)
2779 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2780 lines when AUTH and/or TLS are in use. This is the "received protocol"
2781 field. Exim used to use "asmtp" for authenticated SMTP, without any
2782 indication (in the protocol name) for TLS use. Now it follows the RFC and
2783 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2784 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2785 names appear in log lines as well as in Received: header lines.
2787 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2788 extensions, and comparison tests, and to fix some bugs.
2790 14. Changes to the "personal" filter test:
2792 (1) The test was buggy in that it was just doing the equivalent of
2793 "contains" tests on header lines. For example, if a user's address was
2794 anne@some.where, the "personal" test would incorrectly be true for
2796 To: susanne@some.where
2798 This test is now done by extracting each address from the header in turn,
2799 and checking the entire address. Other tests that are part of "personal"
2800 are now done using regular expressions (for example, to check local parts
2801 of addresses in From: header lines).
2803 (2) The list of non-personal local parts in From: addresses has been
2804 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2805 taken from the Sieve specification recommendations.
2807 (3) If the message contains any header line starting with "List-" it is
2808 treated as non-personal.
2810 (4) The test for "circular" in the Subject: header line has been removed
2811 because it now seems ill-conceived.
2813 15. Minor typos in src/EDITME comments corrected.
2815 16. Installed latest exipick from John Jetmore.
2817 17. If headers_add on a router specified a text string that was too long for
2818 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2819 of string_sprintf() is now avoided.
2821 18. $message_body_size was not set (it was always zero) when running the DATA
2822 ACL and the local_scan() function.
2824 19. For the "mail" command in an Exim filter, no default was being set for
2825 the once_repeat time, causing a random time value to be used if "once" was
2826 specified. (If the value happened to be <= 0, no repeat happened.) The
2827 default is now 0s, meaning "never repeat". The "vacation" command was OK
2828 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2829 (I found it when inspecting the code).
2831 20. There is now an overall timeout for performing a callout verification. It
2832 defaults to 4 times the callout timeout, which applies to individual SMTP
2833 commands during the callout. The overall timeout applies when there is more
2834 than one host that can be tried. The timeout is checked before trying the
2835 next host. This prevents very long delays if there are a large number of
2836 hosts and all are timing out (e.g. when the network connections are timing
2837 out). The value of the overall timeout can be changed by specifying an
2838 additional sub-option for "callout", called "maxwait". For example:
2840 verify = sender/callout=5s,maxwait=20s
2842 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2843 to the end before writing, but this should make it even safer).
2845 22. Exim was forgetting that it had advertised PIPELINING for the second and
2846 subsequent messages on an SMTP connection. It was also not resetting its
2847 memory on STARTTLS and an internal HELO.
2849 23. When Exim logs an SMTP synchronization error within a session, it now
2850 records whether PIPELINING has been advertised or not.
2852 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2853 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2854 rather than long int.
2856 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2858 26. Added the never_mail option to autoreply.
2864 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2865 crash if the getsockname() call failed; this can happen if a connection is
2866 closed very soon after it is established. The problem was simply in the
2867 order in which certain operations were done, causing Exim to try to write
2868 to the SMTP stream before it had set up the file descriptor. The bug has
2869 been fixed by making things happen in the correct order.
2875 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2876 before the connection was closed, thus losing the rejection response.
2878 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2879 some early Solaris releases, but causes trouble in current releases where
2880 socklen_t is defined.
2882 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2885 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2888 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2889 that releasing the top of it at the end releases what was used for sub-
2890 expansions (unless the block got too big). However, discard this block if
2891 the first thing is a variable or header, so that we can use its block when
2892 it is dynamic (useful for very large $message_headers, for example).
2894 6. Lookups now cache *every* query, not just the most recent. A new, separate
2895 store pool is used for this. It can be recovered when all lookup caches are
2896 flushed. Lookups now release memory at the end of their result strings.
2897 This has involved some general refactoring of the lookup sources.
2899 7. Some code has been added to the store_xxx() functions to reduce the amount
2900 of flapping under certain conditions.
2902 8. log_incoming_interface used to affect only the <= reception log lines. Now
2903 it causes the local interface and port to be added to several more SMTP log
2904 lines, for example "SMTP connection from", and rejection lines.
2906 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2908 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2910 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2911 could be overwritten at the end of the current message (or the start of a
2912 new message if it was set in a HELO ACL). The value is now preserved for
2913 the duration of the SMTP connection.
2915 12. If a transport had a headers_rewrite setting, and a matching header line
2916 contained an unqualified address, that address was qualified, even if it
2917 did not match any rewriting rules. The underlying bug was that the values
2918 of the flags that permit the existence of unqualified sender and recipient
2919 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2920 non-local messages, and by -bnq for local messages) were not being
2921 preserved with the message after it was received.
2923 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2924 "next input=" as part of the text comprising the host identity instead of
2925 the correct text. The code was using the same buffer for two different
2926 strings. However, depending on which order the printing function evaluated
2927 its arguments, the bug did not always show up. Under Linux, for example, my
2928 test suite worked just fine.
2930 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2931 doesn't work with some older versions of Perl. It has been changed to "my",
2932 which in any case is probably the better facility to use.
2934 15. A really picky compiler found some instances of statements for creating
2935 error messages that either had too many or two few arguments for the format
2938 16. The size of the buffer for calls to the DNS resolver has been increased
2939 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2940 for addresses that have a lot of PTR records. This alleviates a problem; it
2941 does not fully solve it.
2943 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2944 buffer now truncates the list and logs the incident, which is the same
2945 action as happens when Exim is looking up a host name and its aliases.
2946 Previously in this situation something unpredictable would happen;
2947 sometimes it was "internal error: store_reset failed".
2949 18. If a server dropped the connection unexpectedly when an Exim client was
2950 using GnuTLS and trying to read a response, the client delivery process
2951 crashed while trying to generate an error log message.
2953 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2954 string, for example:
2956 warn message = H1: something\nH2: something
2958 the text was added as a single header line from Exim's point of view
2959 though it ended up OK in the delivered message. However, searching for the
2960 second and subsequent header lines using $h_h2: did not work. This has been
2961 fixed. Similarly, if a system filter added multiple headers in this way,
2962 the routers could not see them.
2964 20. Expanded the error message when iplsearch is called with an invalid key to
2965 suggest using net-iplsearch in a host list.
2967 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2968 ACLs are no longer actually imposed (and a message to that effect is
2971 22. If a "gecos" field in a passwd entry contained escaped characters, in
2972 particular, if it contained a \" sequence, Exim got it wrong when building
2973 a From: or a Sender: header from that name. A second bug also caused
2974 incorrect handling when an unquoted " was present following a character
2975 that needed quoting.
2977 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2978 was not being matched caselessly.
2980 24. Arranged for all hyphens in the exim.8 source to be escaped with
2983 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2984 properly. Recipient callout cache records were still being keyed to include
2985 the sender, even when use_sender was set false. This led to far more
2986 callouts that were necessary. The sender is no longer included in the key
2987 when use_sender is false.
2989 26. Added "control = submission" modifier to ACLs.
2991 27. Added the ${base62d: operator to decode base 62 numbers.
2993 28. dnsdb lookups can now access SRV records.
2995 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2996 the configuration file.
2998 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2999 (-v) mode. This makes the output for a verbose queue run more intelligible.
3001 31. Added a use_postmaster feature to recipient callouts.
3003 32. Added the $body_zerocount variable, containing the number of binary zero
3004 bytes in the message body.
3006 33. The time of last modification of the "new" subdirectory is now used as the
3007 "mailbox time last read" when there is a quota error for a maildir
3010 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3012 35. Added +ignore_unknown as a special item in host lists.
3014 36. Code for decoding IPv6 addresses in host lists is now included, even if
3015 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3016 address was recognized as an IP address, but was then not correctly decoded
3017 into binary, causing unexpected and incorrect effects when compared with
3024 1. Very minor rewording of debugging text in manualroute to say "list of
3025 hosts" instead of "hostlist".
3027 2. If verify=header_syntax was set, and a header line with an unqualified
3028 address (no domain) and a large number of spaces between the end of the
3029 name and the colon was received, the reception process suffered a buffer
3030 overflow, and (when I tested it) crashed. This was caused by some obsolete
3031 code that should have been removed. The fix is to remove it!
3033 3. When running in the test harness, delay a bit after writing a bounce
3034 message to get a bit more predictability in the log output.
3036 4. Added a call to search_tidyup() just before forking a reception process. In
3037 theory, someone could use a lookup in the expansion of smtp_accept_max_
3038 per_host which, without the tidyup, could leave open a database connection.
3040 5. Added the variables $recipient_data and $sender_data which get set from a
3041 lookup success in an ACL "recipients" or "senders" condition, or a router
3042 "senders" option, similar to $domain_data and $local_part_data.
3044 6. Moved the writing of debug_print from before to after the "senders" test
3047 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3048 problems for message scanning, either using a data ACL, or using
3049 local_scan() because the Received: header was not generated till after they
3050 were called (in order to set the time as the time of reception completion).
3051 I have revised the way this works. The header is now generated after the
3052 body is received, but before the ACL or local_scan() are called. After they
3053 are run, the timestamp in the header is updated.
3059 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3060 before starting a queue runner without re-exec. This happened only when
3061 deliver_drop_privilege was set or when the Exim user was set to root. The
3062 effect of the bug was that timeouts during subsequent deliveries caused
3063 crashes instead of being properly handled. The handler is now left at its
3064 default (and expected) setting.
3066 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3067 message, again when deliver_drop_privilege is set or Exim is run as root.
3068 The bug described in (1) was not present in this case, but the tidying up
3069 of the other signals was missing. I have made the two cases consistent.
3071 3. The ignore_target_hosts setting on a manualroute router was being ignored
3072 for hosts that were looked up using the /MX notation.
3074 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3077 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3078 operated on the sender address. After changing the $sender_address to <>
3079 for the sender address verify, Exim was re-instated it as the original
3080 (before rewriting) address, but remembering that it had rewritten it, so it
3081 wasn't rewriting it again. This bug also had the effect of breaking the
3082 sender address verification caching when the sender address was rewritten.
3084 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3085 This has been changed so that if the ip literal address matches
3086 ignore_target_hosts, the router declines.
3088 7. Added expansion conditions match_domain, match_address, and match_local_
3089 part (NOT match_host).
3091 8. The placeholder for the Received: header didn't have a length field set.
3093 9. Added code to Exim itself and to exim_lock to test for a specific race
3094 condition that could lead to file corruption when using MBX delivery. The
3095 issue is with the lockfile that is created in /tmp. If this file is removed
3096 after a process has opened it but before that process has acquired a lock,
3097 there is the potential for a second process to recreate the file and also
3098 acquire a lock. This could lead to two Exim processes writing to the file
3099 at the same time. The added code performs the same test as UW imapd; it
3100 checks after acquiring the lock that its file descriptor still refers to
3101 the same named file.
3103 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3104 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3107 11. Added the smtp_active_hostname option. If used, this will typically be made
3108 to depend on the incoming interface address. Because $interface_address is
3109 not set up until the daemon has forked a reception process, error responses
3110 that can happen earlier (such as "too many connections") no longer contain
3113 12. If an expansion in a condition on a "warn" statement fails because a lookup
3114 defers, the "warn" statement is abandoned, and the next ACL statement is
3115 processed. Previously this caused the whole ACL to be aborted.
3117 13. Added the iplsearch lookup type.
3119 14. Added ident_timeout as a log selector.
3121 15. Added tls_certificate_verified as a log selector.
3123 16. Added a global option tls_require_ciphers (compare the smtp transport
3124 option of the same name). This controls incoming TLS connections.
3126 17. I finally figured out how to make tls_require_ciphers do a similar thing
3127 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3128 before starting the TLS session.
3130 18. Tabs are now shown as \t in -bP output.
3132 19. If the log selector return_path_on_delivery was set, Exim crashed when
3133 bouncing a message because it had too many Received: header lines.
3135 20. If two routers both had headers_remove settings, and the first one included
3136 a superfluous trailing colon, the final name in the first list and the
3137 first name in the second list were incorrectly joined into one item (with a
3138 colon in the middle).
3144 1. Added -C and -D options to the exinext utility, mainly to make it easier
3145 to include in the automated testing, but these could be helpful when
3146 multiple configurations are in use.
3148 2. The exinext utility was not formatting the output nicely when there was
3149 an alternate port involved in the retry record key, nor when there was a
3150 message id as well (for retries that were specific to a specific message
3151 and a specific host). It was also confused by IPv6 addresses, because of
3152 the additional colons they contain. I have fixed the IPv4 problem, and
3153 patched it up to do a reasonable job for IPv6.
3155 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3156 delivery, the log line now contains "pipelined" if PIPELINING was used.
3158 4. An SMTP transport process used to panic and die if the bind() call to set
3159 an explicit outgoing interface failed. This has been changed; it is now
3160 treated in the same way as a connect() failure.
3162 5. A reference to $sender_host_name in the part of a conditional expansion
3163 that was being skipped was still causing a DNS lookup. This no longer
3166 6. The def: expansion condition was not recognizing references to header lines
3167 that used bh_ and bheader_.
3169 7. Added the _cache feature to named lists.
3171 8. The code for checking quota_filecount in the appendfile transport was
3172 allowing one more file than it should have been.
3174 9. For compatibility with Sendmail, the command line option
3182 and sets the incoming protocol and host name (for trusted callers). The
3183 host name and its colon can be omitted when only the protocol is to be set.
3184 Note the Exim already has two private options, -pd and -ps, that refer to
3185 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3186 "s", but I don't think that's a major issue.
3188 10. A number of refactoring changes to the code, none of which should affect
3191 (a) The number of logging options was getting close to filling up the
3192 32-bit word that was used as a bit map. I have split them into two classes:
3193 those that are passed in the argument to log_write(), and those that are
3194 only ever tested independently outside of that function. These are now in
3195 separate 32-bit words, so there is plenty of room for expansion again.
3196 There is no change in the user interface or the logging behaviour.
3198 (b) When building, for example, log lines, the code previously used a
3199 macro that called string_cat() twice, in order to add two strings. This is
3200 not really sufficiently general. Furthermore, there was one instance where
3201 it was actually wrong because one of the argument was used twice, and in
3202 one call a function was used. (As it happened, calling the function twice
3203 did not affect the overall behaviour.) The macro has been replaced by a
3204 function that can join an arbitrary number of extra strings onto a growing
3207 (c) The code for expansion conditions now uses a table and a binary chop
3208 instead of a serial search (which was left over from when there were very
3209 few conditions). Also, it now recognizes conditions like "pam" even when
3210 the relevant support is not compiled in: a suitably worded error message is
3211 given if an attempt is made to use such a condition.
3213 11. Added ${time_interval:xxxxx}.
3215 12. A bug was causing one of the ddress fields not to be passed back correctly
3216 from remote delivery subprocesses. The field in question was not being
3217 subsequently used, so this caused to problems in practice.
3219 13. Added new log selectors queue_time and deliver_time.
3221 14. Might have fixed a bug in maildirsizefile handling that threw up
3222 "unexpected character" debug warnings, and recalculated the data
3223 unnecessarily. In any case, I expanded the warning message to give more
3226 15. Added the message "Restricted characters in address" to the statements in
3227 the default ACL that block characters like @ and % in local parts.
3229 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3230 Three changes have been made:
3232 (a) There was a serious bug; a negative response to MAIL caused the whole
3233 recipient domain to be cached as invalid, thereby blocking all messages
3234 to all local parts at the same domain, from all senders. This bug has
3235 been fixed. The domain is no longer cached after a negative response to
3236 MAIL if the sender used is not empty.
3238 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3241 (c) A new callout option, "use_sender" has been added for people who want
3242 the modified behaviour.
3248 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3251 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3252 indeed breaks things for older releases.
3254 3. Added additional logging to the case where there is a problem reading data
3255 from a filter that is running in a subprocess using a pipe, in order to
3256 try to track down a specific problem.
3258 4. Testing facility fudge: when running in the test harness and attempting
3259 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3260 getting "No route to host". Convert this to a timeout.
3262 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3265 6. Some OS don't have socklen_t but use size_t instead. This affects the
3266 fifth argument of getsockopt() amongst other things. This is now
3267 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3268 can be set for individual OS. I have set it for SunOS5, OSF1, and
3269 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3270 some earlier ones do not.
3272 7. Change 4.30/15 was not doing the test caselessly.
3274 8. The standard form for an IPv6 address literal was being rejected by address
3275 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3276 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3277 this, as well as the form without the "IPv6" on the front (but only when
3278 address literals are enabled, of course).
3280 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3282 10. Exim crashed if a message with an empty sender address specified by -f
3283 encountered a router with an errors_to setting. This could be provoked only
3284 by a command such as
3288 where an empty string was supplied; "<>" did not hit this bug.
3290 11. Installed PCRE release 4.5.
3292 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3293 remained set. It is now erased.
3295 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3296 times from message ids (which are base 36 rather than the normal 62).
3298 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3299 were being counted as actual protocol errors, and logged if the log
3300 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3301 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3302 and DATA following a set of rejected RCPTs do not count as protocol errors.
3303 In other words, Exim assumes they were pipelined, though this may not
3304 actually be the case. Of course, in all cases the client gets an
3305 appropriate error code.
3307 15. If a lookup fails in an ACL condition, a message about the failure may
3308 be available; it is used if testing the ACL cannot continue, because most
3309 such messages specify what the cause of the deferral is. However, some
3310 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3311 that caused an old message to be retained and used if a later statement
3312 caused a defer, replacing the real cause of the deferral.
3314 16. If an IP address had so many PTR records that the DNS lookup buffer
3315 was not large enough to hold them, Exim could crash while trying to process
3316 the truncated data. It now detects and logs this case.
3318 17. Further to 4.21/58, another change has been made: if (and only if) the
3319 first line of a message (the first header line) ends with CRLF, a bare LF
3320 in a subsequent header line has a space inserted after it, so as not to
3321 terminate the header.
3323 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3324 unnecessarily, used atoi() instead of strtol(), and didn't check the
3325 termination when getting file sizes from file names by regex.
3327 19. Completely re-implemented the support for maildirsize files, in the light
3328 of a number of problems with the previous contributed implementation
3329 (4.30/29). In particular:
3331 . If the quota is zero, the maildirsize file is maintained, but no quota is
3334 . If the maildir directory does not exist, it is created before any attempt
3335 to write a maildirsize file.
3337 . The quota value in the file is just a cache; if the quota is changed in
3338 the transport, the new value overrides.
3340 . A regular expression is available for excluding directories from the
3343 20. The autoreply transport checks the characters in options that define the
3344 message's headers; it allows continued headers, but it was checking with
3345 isspace() after an embedded newline instead of explicitly looking for a
3348 21. If all the "regular" hosts to which an address was routed had passed their
3349 expiry times, and had not reached their retry times, the address was
3350 bounced, even if fallback hosts were defined. Now Exim should go on to try
3353 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3354 equivalent code in the SMTP transport. Some hosts send humungous responses
3355 to HELO/EHLO, more than 1024 it seems.
3357 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3358 gives compiler warnings in some environments. I've now done it "properly",
3361 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3362 (because of problems with the built-in one) was declared to return uschar *
3363 instead of char *, causing compiler failure.
3365 25. Fixed a file descriptor leak when processing alias/forward files.
3367 26. Fixed a minor format string issue in dbfn.c.
3369 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3371 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3372 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3375 29. When a sender address is verified, it is cached, to save repeating the test
3376 when there is more than one recipient in a message. However, when the
3377 verification involves a callout, it is possible for different callout
3378 options to be set for different recipients. It is too complicated to keep
3379 track of this in the cache, so now Exim always runs a verification when a
3380 callout is required, relying on the callout cache for the optimization.
3381 The overhead is duplication of the address routing, but this should not be
3384 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3385 to be verified with callout=postmaster, and the main callout worked but the
3386 postmaster check failed, the verification correctly failed. However, if a
3387 subsequent RCPT command asked for sender verification *without* the
3388 postmaster check, incorrect caching caused this verification also to fail,
3391 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3392 it was not caching the DNS options (qualify_single, search_parents) that
3393 were used when the lookup failed. A subsequent lookup with different
3394 options therefore always gave the same answer, though there were cases
3395 where it should not have. (Example: a "domains = !$mx_any" option on a
3396 dnslookup router: the "domains" option is always processed without any
3397 widening, but the router might have qualify_single set.) Now Exim uses the
3398 cached value only when the same options are set.
3400 32. Added John Jetmore's "exipick" utility to the distribution.
3402 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3403 than a timeout (e.g. a certificate is required, and is not provided), an
3404 Exim server now closes the connection immediately. Previously it waited for
3405 the client to close - but if the client is SSL, it seems that they each
3406 wait for each other, leading to a delay before one of them times out.
3408 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3409 maintained 0.8.x compatibility because I don't think many are using it, and
3410 it is clearly obsolete.
3412 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3415 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3416 client certificate was expired. A simple patch fixes this, though I don't
3417 understand the full logic of why the verify callback is called multiple
3420 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3421 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3422 which causes problems with some clients (such as the Certicom SSL Plus
3423 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3424 disables the coutermeasure allowing Eudora to connect."
3426 38. Exim was not checking that a write() to a log file succeeded. This could
3427 lead to Bad Things if a log got too big, in particular if it hit a file
3428 size limit. Exim now panics and dies if it cannot write to a log file, just
3429 as it does if it cannot open a log file.
3431 39. Modified OS/Makefile-Linux so that it now contains
3433 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3435 The two -D definitions ensure that Exim is compiled with large file
3436 support, which makes it possible to handle log files that are bigger than
3439 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3440 instance) a domain was checked against a named list that involved a lookup,
3441 causing $domain_data to be set, then another domain was checked against the
3442 same list, then the first domain was re-checked, the value of $domain_data
3443 after the final check could be wrong. In particular, if the second check
3444 failed, it could be set empty. This bug probably also applied to
3447 41. The strip_trailing_dot option was not being applied to the address given
3448 with the -f command-line option.
3450 42. The code for reading a message's header from the spool was incrementing
3451 $received_count, but never initializing it. This meant that the value was
3452 incorrect (doubled) while delivering a message in the same process in which
3453 it was received. In the most common configuration of Exim, this never
3454 happens - a fresh exec is done - but it can happen when
3455 deliver_drop_privilege is set.
3457 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3458 it now includes up to 150 characters of the unexpected data in the log
3461 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3462 and building data strings. The size of both of these buffers was 10 000
3463 bytes - far larger than anybody would *ever* want, thought I. Needless to
3464 say, somebody hit the limit. I have increased the maximum line length to
3465 20 000 and the maximum data length of concatenated lines to 100 000. I have
3466 also fixed two bugs, because there was no checking on these buffers. Tsk,
3467 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3468 buffer is too small.
3470 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3471 lsearch lookups. Now it does.
3473 46. When parsing a route_list item in a manualroute router, a fixed-length
3474 buffer was used for the list of hosts. I made this 1024 bytes long,
3475 thinking that nobody would ever have a list of hosts that long. Wrong.
3476 Somebody had a whole pile of complicated expansion conditions, and the
3477 string was silently truncated, leading to an expansion error. It turns out
3478 that it is easier to change to an unlimited length (owing to other changes
3479 that have happened since this code was originally written) than to build
3480 structure for giving a limitation error. The length of the item that
3481 expands into the list of hosts is now unlimited.
3483 47. The lsearch lookup could not handle data where the length of text line was
3484 more than 4095 characters. Such lines were truncated, leading to shortened
3485 data being returned. It should now handle lines of any length.
3487 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3488 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3491 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3492 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3493 no longer happens in comment lines. A list of replacements is now placed
3494 at the head of all of the source files, except those whose only change is
3495 to replace PERL_COMMAND in the very first #! line.
3497 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3498 messages on the queue, with a bottom-up merge sort, using code contributed
3499 by Michael Haardt. This should make operations like -bp somewhat faster on
3500 large queues. It won't affect queue runners, except when queue_run_in_order
3503 51. Installed eximstats 1.31 in the distribution.
3505 52. Added support for SRV lookups to the dnslookup router.
3507 53. If an ACL referred to $message_body or $message_body_end, the value was not
3508 reset for any messages that followed in the same SMTP session.
3510 54. The store-handling optimization for building very long strings was not
3511 differentiating between the different store pools. I don't think this
3512 actually made any difference in practice, but I've tidied it.
3514 55. While running the routers to verify a sender address, $sender_address
3515 was still set to the sender address. This is wrong, because when routing to
3516 send a bounce to the sender, it would be empty. Therefore, I have changed
3517 it so that, while verifying a sender address, $sender_address is set to <>.
3518 (There is no change to what happens when verifying a recipient address.)
3520 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3521 target A or AAAA records (if not already returned) without resetting the
3522 qualify_single or search_parents options of the DNS resolver. These are
3523 inappropriate in this case because the targets of MX and SRV records must
3524 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3525 target that, when qualified, matched something in the local domain. These
3526 two options are now turned off when doing these lookups.
3528 57. It seems that at least some releases of Reiserfs (which does not have the
3529 concept of a fixed number of inodes) returns zero and not -1 for the
3530 number of available inodes. This interacted badly with check_spool_inodes,
3531 which assumed that -1 was the "no such thing" setting. What I have done is
3532 to check that the total number of inodes is greater than zero before doing
3533 the test of how many are available.
3535 58. When a "warn" ACL statement has a log_message modifier, the message is
3536 remembered, and not repeated. This is to avoid a lot of repetition when a
3537 message has many recipients that cause the same warning to be written.
3538 Howewer, Exim was preserving the list of already written lines for an
3539 entire SMTP session, which doesn't seem right. The memory is now reset if a
3540 new message is started.
3542 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3543 debugging output unless log_rewrite was also set.
3545 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3546 of (int)(handle) when we know that handle contains (void *)(-1).
3548 61. The Exim daemon panic-logs an error return when it closes the incoming
3549 connection. However "connection reset by peer" seems to be common, and
3550 isn't really an error worthy of noting specially, so that particular error
3553 62. When Exim is trying to find all the local interfaces, it used to panic and
3554 die if the ioctl to get the interface flags failed. However, it seems that
3555 on at least one OS (Solaris 9) it is possible to have an interface that is
3556 included in the list of interfaces, but for which you get a failure error
3557 for this call. This happens when the interface is not "plumbed" into a
3558 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3559 failure of the "get flags" call assumes that the interface is down.
3561 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3562 makes life easier for people who are doing arithmetic on fields extracted
3563 from dates, where you often get leading zeros that should not be
3564 interpreted as octal.
3566 64. Added qualify_domain to the redirect router, to override the global
3569 65. If a pathologically long header line contained very many addresses (the
3570 report of this problem mentioned 10 000) and each of them was rewritten,
3571 Exim could use up a very large amount of memory. (It kept on making new
3572 copies of the header line as it rewrote, and never released the old ones.)
3573 At the expense of a bit more processing, the header rewriting function has
3574 been changed so that it no longer eats memory in this way.
3576 66. The generation of the Received: header has been moved from the time that a
3577 message starts to be received, to the time that it finishes. The timestamp
3578 in the Received: header should now be very close to that of the <= log
3579 line. There are two side-effects of this change:
3581 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3582 logged header lines no longer include the local Received: line, because
3583 it has not yet been created. The same applies to a copy of the message
3584 that is returned to a non-SMTP sender when a message is rejected.
3586 (b) When a filter file is tested using -bf, no additional Received: header
3587 is added to the test message. After some thought, I decided that this
3590 This change does not affect the value of $received_for. It is still set
3591 after address rewriting, but before local_scan() is called.
3593 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3595 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3596 gave an unhelpful panic error message, and a defer error. I have managed to
3597 change this behaviour so that it now rejects any supplied certificate,
3598 which seems right, as the list of acceptable certificates is empty.
3600 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3601 gave an unhelpful defer error. I have not managed to make this reject any
3602 supplied certificates, but the error message it gives is "no certificate
3603 supplied", which is not helpful.
3605 70. exigrep's output now also includes lines that are not associated with any
3606 message, but which match the given pattern. Implemented by a patch from
3607 Martin Sluka, which also tidied up the Perl a bit.
3609 71. Recipient callout verification, like sender verification, was using <> in
3610 the MAIL FROM command. This isn't really the right thing, since the actual
3611 sender may affect whether the remote host accepts the recipient or not. I
3612 have changed it to use the actual sender in the callout; this means that
3613 the cache record is now keyed on a recipient/sender pair, not just the
3614 recipient address. There doesn't seem to be a real danger of callout loops,
3615 since a callout by the remote host to check the sender would use <>.
3616 [SEE ABOVE: changed after hitting problems.]
3618 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3619 temporary errors. However, in the case of such a code being given after
3620 the end of a data transmission (i.e. after ".") Exim was failing to write
3621 a retry record for the message. (Yes, there was some broken host that was
3622 actually sending 8xx at this point.)
3624 73. An unknown lookup type in a host list could cause Exim to panic-die when
3625 the list was checked. (An example that provoked this was putting <; in the
3626 middle of a list instead of at the start.) If this happened during a DATA
3627 ACL check, a -D file could be left lying around. This kind of configuration
3628 error no longer causes Exim to die; instead it causes a defer errror. The
3629 incident is still logged to the main and panic logs.
3631 74. Buglet left over from Exim 3 conversion. The message "too many messages
3632 in one connection" was written to the rejectlog but not the mainlog, except
3633 when address rewriting (yes!) was being logged.
3635 75. Added write_rejectlog option.
3637 76. When a system filter was run not as root (that is, when system_filter_user
3638 was set), the values of the $n variables were not being returned to the
3639 main process; thus, they were not subsequently available in the $sn
3642 77. Added +return_path_on_delivery log selector.
3644 78. A connection timeout was being treated differently from recipients deferred
3645 when testing hosts_max_try with a message that was older than the host's
3646 retry timeout. (The host should not be counted, thus allowing all hosts to
3647 be tried at least once before bouncing.) This may have been the cause of an
3648 occasionally reported bug whereby a message would remain on the queue
3649 longer than the retry timeout, but would be bounced if a delivery was
3650 forced. I say "may" because I never totally pinned down the problem;
3651 setting up timeout/retry tests is difficult. See also the next item.
3653 79. The ultimate address timeout was not being applied to errors that involved
3654 a combination of host plus message (for example, a timeout on a MAIL
3655 command). When an address resolved to a number of possible hosts, and they
3656 were not all tried for each delivery (e.g. because of hosts_max_try), a
3657 message could remain on the queue longer than the retry timeout.
3659 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3662 81. Fixed an obscure SMTP outgoing bug which required at least the following
3663 conditions: (a) there was another message waiting for the same server;
3664 (b) the server returned 5xx to all RCPT commands in the first message so
3665 that the message was not completed; (c) the server dropped the connection
3666 or gave a negative response to the RSET that Exim sends to abort the
3667 transaction. The observed case was a dropped connection after DATA that had
3668 been sent in pipelining mode. That is, the server had advertised PIPELINING
3669 but was not implementing it correctly. The effect of the bug was incorrect
3670 behaviour, such as trying another host, and this could lead to a crash.
3676 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3677 and daemon.c were passed as pointers to ints; they should have been
3678 pointers to socklen_t variables (which are typically unsigned ints).
3680 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3683 3. Fixed a really odd bug that affected only the testing scheme; patching a
3684 certain fixed string in the binary changed the value of another string that
3685 happened to be identical to the end of the original first string.
3687 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3688 name", it returns that address as the IP address. On some operating
3689 systems (e.g. Solaris), it also passes back the IP address string as the
3690 "host name". However, on others (e.g. Linux), it passes back an empty
3691 string. Exim wasn't checking for this, and was changing the host name to an
3692 empty string, assuming it had been canonicized.
3694 5. Although rare, it is permitted to have more than one PTR record for a given
3695 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3696 all the names associated with an address, because they do in Solaris.
3697 However, it seems that they do not in Linux for data that comes from the
3698 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3699 I found this out when I moved to a new Linux workstation and tried to run
3700 the Exim test suite.
3702 To get round this problem I have changed the code so that it now does its
3703 own call to the DNS to look up PTR records when searching for a host name.
3704 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3705 addresses that are only in /etc/hosts are still found.
3707 This behaviour is, however, controlled by an option called host_lookup_
3708 order, which defaults to "bydns:byaddr". If people want to use the other
3709 order, or indeed, just use one or the other means of lookup, they can
3710 specify it in this variable.
3712 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3713 some operating systems, this comes back from gethostbyaddr() as an empty
3714 string, and this is what Exim used to test for. However, it seems that in
3715 other systems, "." is yielded. Exim now tests for this case too.
3717 7. The values of check_spool_space and check_log_space are now held internally
3718 as a number of kilobytes instead of an absolute number of bytes. If a
3719 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3720 kilobyte. This means that much larger values can be stored.
3722 8. Exim monitor: an attempt to get the action menu when not actually pointing
3723 at a message produces an empty menu entitled "No message selected". This
3724 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3725 no entries in it ("Shell widget menu has zero width and/or height"). So I
3726 have added a single, blank menu entry in this case.
3728 9. Added ${quote_local_part.
3730 10. MIME decoding is now applied to the contents of Subject: header lines when
3733 11. Now that a reference to $sender_host_address automatically causes a reverse
3734 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3735 host lookup before query-style lookups in lists that might use this
3736 variable. This has therefore been abolished, and the "net-" prefix is no
3737 longer necessary for query-style lookups.
3739 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3740 have been a typo for LFLAGS, so it has been changed.
3742 13. The install script calls Exim with "-C /dev/null" in order to find the
3743 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3744 to be output. Howeve, since Exim outputs its version number before the
3745 error, it didn't break the script. It just looked ugly. I fixed this by
3746 always allowing "-C /dev/null" if the caller is root.
3748 14. Ignore overlarge ACL variable number when reading spool file - insurance
3749 against a later release with more variables having written the file.
3751 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3752 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3753 this, as well as the form without the "IPv6" on the front.
3755 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3756 OS/Makefile-Darwin file.
3758 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3759 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3761 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3762 variable. However, some OS use ints for inodes. I've added cast to long int
3763 to get rid of the compiler warning.
3765 19. I had forgotten to lock out "/../" in configuration file names when
3766 ALT_CONFIG_PREFIX was set.
3768 20. Routers used for verification do not need to specify transports. However,
3769 if such a router generated a host list, and callout was configured, Exim
3770 crashed, because it could not find a port number from the (non-existent)
3771 transport. It now assumes port 25 in this circumstance.
3773 21. Added the -t option to exigrep.
3775 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3776 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3777 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3778 them is set, LOOKUP_LSEARCH is forced.
3780 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3782 24. Added sender and host information to the "rejected by local_scan()" log
3783 line; previously there was no indication of these.
3785 25. Added .include_if_exists.
3787 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3788 Linux. It turns out that not all file systems support this. Apparently some
3789 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3790 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3791 not supported on the file descriptor, is now ignored when Exim is trying to
3792 sync a directory. This applies only to Linux.
3794 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3796 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3797 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3800 29. Added support for maildirsize files from supplied patch (modified a bit).
3802 30. The use of :fail: followed by an empty string could lead Exim to respond to
3803 sender verification failures with (e.g.):
3805 550 Verification failed for <xxx>
3806 550 Sender verify failed
3808 where the first response line was missing the '-' that indicates it is not
3809 the final line of the response.
3811 31. The loop for finding the name of the user that called Exim had a hardwired
3812 limit of 10; it now uses the value of finduser_retries, which is used for
3813 all other user lookups.
3815 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3818 33. Exim was neglecting to zero errno before one call of strtol() when
3819 expanding a string and expecting an integer value. On some systems this
3820 resulted in spurious "integer overflow" errors. Also, it was casting the
3821 result into an int without checking.
3823 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3824 did not work. The code looks as if it has *never* worked, though it appears
3825 to have been documented since at least releast 1.62. I have made it work.
3827 35. The "timeout_DNS" error in retry rules, also documented since at least
3828 1.62, also never worked. As it isn't clear exactly what this means, and
3829 clearly it isn't a major issue, I have abolished the feature by treating it
3830 as "timeout", and writing a warning to the main and panic logs.
3832 36. The display of retry rules for -brt wasn't always showing the error code
3835 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3836 timeout_connect_A, timeout_connect_MX.
3838 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3839 to the empty sender.
3841 39. The daemon was not analysing the content of -oX till after it had closed
3842 stderr and disconnected from the controlling terminal. This meant that any
3843 syntax errors were only noted on the panic log, and the return code from
3844 the command was 0. By re-arranging the code a little, I've made the
3845 decoding happen first, so such errors now appear on stderr, and the return
3846 code is 1. However, the actual setting up of the sockets still happens in
3847 the disconnected process, so errors there are still only recorded on the
3850 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3851 connections (as happens on some IP stacks) was logged at start up time as
3852 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3853 it from "IPv6 and IPv4", which means that two separate sockets are being
3856 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3857 says whether AF_INET or AF_INET6 was passed as an argument.
3859 42. Exiwhat output was messed up when time zones were included in log
3862 43. Exiwhat now gives more information about the daemon's listening ports,
3863 and whether -tls-on-connect was used.
3865 44. The "port" option of the smtp transport is now expanded.
3867 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3868 silently ignored. Now an error message is written to the main and panic
3871 46. There's a new ACL modifier called "logwrite" which writes to a log file
3872 as soon as it is encountered.
3874 47. Added $local_user_uid and $local_user_gid at routing time.
3876 48. Exim crashed when trying to verify a sender address that was being
3879 49. Exim was recognizing only a space character after ".include". It now also
3880 recognizes a tab character.
3882 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3883 extracting the relevant information from the specification. The man page no
3884 longer contains scrambled data for the -d option, and I've added a section
3885 at the front about calling Exim under different names.
3887 51. Added "extra_headers" argument to the "mail" command in filter files.
3889 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3892 53. Installed eximstats 1.29.
3894 54. Added transport_filter_timeout as a generic transport option.
3896 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3897 Cc: header lines. This was required by RFC 822, but it not required by RFC
3900 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3901 incoming messages that did not have them. Now it does so only if the
3902 message originates locally, that is, if there is no associated remote host
3903 address. When Resent- header lines are present, this applies to the Resent-
3904 lines rather than the non-Resent- lines.
3906 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3907 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3909 58. Messages for configuration errors now include the name of the main
3910 configuration file - useful now that there may be more than one file in a
3911 list (.included file names were always shown).
3913 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3914 for those rare installations that do not start the daemon as root or run it
3915 setuid root. I've cut out the call to initgroups() if the daemon is not
3918 60. The Exim user and group can now be bound into the binary as text strings
3919 that are looked up at the start of Exim's processing.
3921 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3923 62. Added $mailstore_basename variable.
3925 63. Installed patch to sieve.c from Michael Haardt.
3927 64. When Exim failed to open the panic log after failing to open the main log,
3928 the original message it was trying to log was written to stderr and debug
3929 output, but if they were not available (the usual case in production), it
3930 was lost. Now it is written to syslog before the two lines that record the
3931 failures to open the logs.
3933 65. Users' Exim filters run in subprocesses under the user's uid. It is
3934 possible for a "deliver" command or an alias in a "personal" command to
3935 provoke an address rewrite. If logging of address rewriting is configured,
3936 this fails because the process is not running as root or exim. There may be
3937 a better way of dealing with this, but for the moment (because 4.30 needs
3938 to be released), I have disabled address rewrite logging when running a
3939 filter in a non-root, non-exim process.
3945 1. The buildconfig auxiliary program wasn't quoting the value set for
3946 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3947 not defined. This bug was masked in 4.22 by the effect that was fixed in
3950 2. Some messages that were rejected after a message id was allocated were
3951 shown as "incomplete" by exigrep. It no longer does this for messages that
3952 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3954 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3955 have allow_domain_literals set, the ID did not get logged in the <= line.
3956 Domain literals are now always recognized in Message-ID: header lines.
3958 4. The first argument for a ${extract expansion item is the key name or field
3959 number. Leading and trailing spaces in this item were not being ignored,
3960 causing some misleading effects.
3962 5. When deliver_drop_privilege was set, single queue runner processes started
3963 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3964 same command in the process it spins off) were not dropping privilege.
3966 6. When the daemon running as "exim" started a queue runner, it always
3967 re-executed Exim in the spun-off process. This is a waste of effort when
3968 deliver_drop_privilege is set. The new process now just calls the
3969 queue-runner function directly.
3975 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3978 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3979 ignored. Though the use of -oP was forcing the writing of a pid file, it
3980 was always written to the default place.
3982 3. If the message "no IP address found for host xxxx" is generated during
3983 incoming verification, it is now followed by identification of the incoming
3984 connection (so you can more easily find what provoked it).
3986 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3988 5. Added some features to "harden" Exim a bit more against certain attacks:
3990 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3991 be put in Local/Makefile. This is like the never_users runtime option,
3992 but it cannot be overridden. The default setting is "root".
3994 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3995 prefix string with which any file named in a -C command line option
3998 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3999 is retained for -C and -D only if the caller of Exim is root. Without
4000 it, the exim user may also use -C and -D and retain privilege.
4002 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4003 command line option is disabled.
4005 6. Macro names set by the -D option must start with an upper case letter, just
4006 like macro names defined in the configuration file.
4008 7. Added "dereference=" facility to LDAP.
4010 8. Two instances of the typo "uknown" in the source files are fixed.
4012 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4013 the Configure-Makefile script screwed up while processing it.
4015 10. Incorporated PCRE 4.4.
4017 11. The SMTP synchronization check was not operating right at the start of an
4018 SMTP session. For example, it could not catch a HELO sent before the client
4019 waited for the greeting. There is now a check for outstanding input at the
4020 point when the greeting is written. Because of the duplex, asynchronous
4021 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4022 way, but not yet received, when the check is performed.
4024 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4025 on TCP/IP sockets, because this apparently causes some broken clients to
4028 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4029 unchanged) from the Cygwin maintainer.
4031 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4032 was supported instead of testing for mbx. Effectively a typo.
4034 15. The spa authenticator server code was not checking that the input it
4035 received was valid base64.
4037 16. The debug output line for the "set" modifier in ACLs was not showing the
4038 name of the variable that was being set.
4040 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4042 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4043 lookup on its own. Something else was needed to trigger the lookup. For
4044 example, a match in host_lookup or the need for a host name in a host list.
4045 Now, if $sender_host_name is referenced and the host name has not yet been
4046 looked up, a lookup is performed. If the lookup fails, the variable remains
4047 empty, and $host_lookup_failed is set to "1".
4049 19. Added "eqi" as a case-independent comparison operator.
4051 20. The saslauthd authentication condition could segfault if neither service
4052 nor realm was specified.
4054 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4055 error message that was logged was misleading, and incoming SMTP
4056 connections were dropped. The message is now more accurate, and temporary
4057 errors are given to SMTP connections.
4059 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4060 (except RSET) with a 421 error, until QUIT is received. However, it was
4061 failing to send a response to QUIT.
4063 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4064 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4065 ACL is now run after the helo_try_verify_hosts code.
4067 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4068 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4069 case-independent, so other case variants are also recognized). Apparently
4070 some systems use these upper case variants.
4072 25. If more than two messages were waiting for the same host, and a transport
4073 filter was specified for the transport, Exim sent two messages over the
4074 same TCP/IP connection, and then failed with "socket operation on non-
4075 socket" when it tried to send the third.
4077 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4079 27. The extern definition of crypt16() in expand.c was not being excluded when
4080 the OS had its own crypt16() function.
4082 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4083 as a preferred synonym for return_size_limit, both as an option and as an
4086 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4088 30. Changed the default configuration ACL to relax the local part checking rule
4089 for addresses that are not in any local domains. For these addresses,
4090 slashes and pipe symbols are allowed within local parts, but the sequence
4091 /../ is explicitly forbidden.
4093 31. SPA server authentication was not clearing the challenge buffer before
4096 32. log_message in a "warn" ACL statement was writing to the reject log as
4097 well as to the main log, which contradicts the documentation and doesn't
4098 seem right (because no rejection is happening). So I have stopped it.
4100 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4101 However, I am unable to do any testing of this.
4103 34. Fixed an infelicity in the appendfile transport. When checking directories
4104 for a mailbox, to see if any needed to be created, it was accidentally
4105 using path names with one or more superfluous leading slashes; tracing
4106 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4108 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4109 contents are added to the log line that is written for every discarded
4110 recipient. (Previously a log_message setting was ignored.)
4112 36. The ${quote: operator now quotes the string if it is empty.
4114 37. The install script runs exim in order to find its version number. If for
4115 some reason other than non-existence or emptiness, which it checks, it
4116 could not run './exim', it was installing it with an empty version number,
4117 i.e. as "exim-". This error state is now caught, and the installation is
4120 38. An argument was missing from the function that creates an error message
4121 when Exim fails to connect to the socket for saslauthd authentication.
4122 This could cause Exim to crash, or give a corrupted message.
4124 39. Added isip, isip4, and isip6 to ${if conditions.
4126 40. The ACL variables $acl_xx are now saved with the message, and can be
4127 accessed later in routers, transports, and filters.
4129 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4130 strings in the file are not string-expanded.
4132 42. If a MAIL command specified a SIZE value that was too large to fit into an
4133 int variable, the check against message_size_limit failed. Such values are
4134 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4135 day this will have to be increased, but I don't think I want to be around
4136 when emails are that large.
4143 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4144 iconv() is not standard in FreeBSD.
4146 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4147 IPv6 enabled. The observed symptom was a segmentation fault on return from
4148 the function os_common_find_running_interfaces() in src/os.c.
4150 3. In the check_special_case() function in daemon.c I had used "errno" as an
4151 argument name, which causes warnings on some systems. This was basically a
4152 typo, since it was named "eno" in the comments!
4154 4. The code that waits for the clock to tick (at a resolution of some fraction
4155 of a second) so as to ensure message-id uniqueness was always waiting for
4156 at least one whole tick, when it could have waited for less. [This is
4157 almost certainly not relevant at current processor speeds, where it is
4158 unlikely to ever wait at all. But we try to future-proof.]
4160 5. The function that sleeps for a time interval that includes fractions of a
4161 second contained a race. It did not block SIGALRM between setting the
4162 timer, and suspending (a couple of lines later). If the interval was short
4163 and the sigsuspend() was delayed until after it had expired, the suspension
4164 never ended. On busy systems this could lead to processes getting stuck for
4167 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4168 process, before it forks any delivery processes. The open lookup caching
4169 mechanism meant that the open file or database connection was passed into
4170 the delivery process. The problem was that delivery processes always tidy
4171 up cached lookup data. This could cause a problem for the next delivery
4172 process started by the queue runner, because the external queue runner
4173 process does not know about the closure. So the next delivery process
4174 still has data in the lookup cache. In the case of a file lookup, there was
4175 no problem because closing a file descriptor in a subprocess doesn't affect
4176 the parent. However, if the lookup was caching a connection to a database,
4177 the connection was closed, and the second delivery process was likely to
4178 see errors such as "PGSQL: query failed: server closed the connection
4179 unexpectedly". The problem has been fixed by closing all cached lookups
4180 in a queue runner before running a delivery process.
4182 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4183 seem to have the "const" qualifier which it has on other OS. I've
4186 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4187 *of the same type* (client or server) with the same public name that an
4188 error should be diagnosed.
4190 9. When Exim looked up a host name for an IP address, but failed to find the
4191 original IP address when looking up the host name (a safety check), it
4192 output the message "<ip address> does not match any IP for NULL", which was
4193 confusing, to say the least. The bug was that the host name should have
4194 appeared instead of "NULL".
4196 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4197 user that is built into the binary, and the -C or -D options is used, root
4198 privilege is dropped before the configuration file is read. In addition,
4199 logging is switched to stderr instead of the normal log files. If the
4200 configuration then re-defines the Exim user, the unprivileged environment
4201 is probably not what is expected, so Exim logs a panic warning message (but
4204 However, if deliver_drop_privilege is set, the unprivileged state may well
4205 be exactly what is intended, so the warning has been cut out in that case,
4206 and Exim is allowed to try to write to its normal log files.
4212 1. smtp_return_error_details was not giving details for temporary sender
4213 or receiver verification errors.
4215 2. Diagnose a configuration error if two authenticators have the same public
4218 3. Exim used not to create the message log file for a message until the first
4219 delivery attempt. This could be confusing when incoming messages were held
4220 for policy or load reasons. The message log file is now created at the time
4221 the message is received, and an initial "Received" line is written to it.
4223 4. The automatically generated man page for command line options had a minor
4224 bug that caused no ill effects; however, a more serious problem was that
4225 the procedure for building the man page automatically didn't always
4226 operate. Consequently, release 4.20 contains an out-of-date version. This
4227 shouldn't happen again.
4229 5. When building Exim with embedded Perl support, the script that builds the
4230 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4231 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4233 6. The freeze_tell option was not being used for messages that were frozen on
4234 arrival, either by an ACL or by local_scan().
4236 7. Added the smtp_incomplete_transaction log selector.
4238 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4239 was accepting AUTH without a new EHLO.
4241 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4242 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4245 10. Logging of TCP/IP connections (when configured) now happens in the main
4246 daemon process instead of the child process, so that the TCP/IP connection
4247 count is more accurate (but it can never be perfect).
4249 11. The use of "drop" in a nested ACL was not being handled correctly in the
4250 outer ACL. Now, if condition failure induced by the nested "drop" causes
4251 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4252 or "require"), the connection is dropped.
4254 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4255 that yield "discard" can now be used with an "accept" or a "discard" verb,
4256 but an error is generated for any others (because I can't see a useful way
4257 to define what should happen).
4259 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4260 are now processed in the same way as lines in the Exim configuration file.
4261 In particular, continuation lines are supported.
4263 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4265 15. Added -ti meaning -t -i.
4267 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4268 domains, and warn by logging if others are found.
4270 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4271 returned from GIFCONF when Exim is trying to find the list of interfaces on
4272 a host. The code in os.c has been modified to copy each ifreq to an aligned
4273 structure in all cases.
4275 Also, in some cases, the returned ifreq's were being copied to a 'struct
4276 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4277 means the last couple of bytes of an IPv6 address could be chopped if the
4278 ifreq contained only a normal sockaddr (14 bytes storage).
4280 18. Named domain lists were not supported in the hosts_treat_as_local option.
4281 An entry such as +xxxx was not recognized, and was treated as a literal
4284 19. Ensure that header lines added by a DATA ACL are included in the reject log
4285 if the ACL subsequently rejects the message.
4287 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4288 MD5 (which is deprecated).
4290 21. When testing a filter file using -bf, Exim was writing a message when it
4291 took the sender from a "From " line in the message, but it was not doing so
4292 when it took $return_path from a Return-Path: header line. It now does.
4294 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4295 with a valid header line field name (a series of printing characters
4296 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4298 23. Changed "disc" in the source to "disk" to conform to the documentation and
4299 the book and for uniformity.
4301 24. Ignore Sendmail's -Ooption=value command line item.
4303 25. When execve() failed while trying to run a command in a pipe transport,
4304 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4305 could be confused with a return value of 69 from the command itself. This
4306 has been changed to 127, the value the shell returns if it is asked to run
4307 a non-existent command. The wording for the related log line suggests a
4308 non-existent command as the problem.
4310 26. If received_header_text expands to an empty string, do not add a Received:
4311 header line to the message. (Well, it adds a token one on the spool, but
4312 marks it "old" so that it doesn't get used or transmitted.)
4314 27. Installed eximstats 1.28 (addition of -nt option).
4316 28. There was no check for failure on the call to getsockname() in the daemon
4317 code. This can fail if there is a shortage of resources on the system, with
4318 ENOMEM, for example. A temporary error is now given on failure.
4320 29. Contrary to the C standard, it seems that in some environments, the
4321 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4322 program. Exim now does this explicitly; it affects the formatting of
4323 timestamps using strftime().
4325 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4326 complaints. I've now initialized all the variables, to avoid this.
4328 32. Header lines added by a system filter were not being "seen" during
4329 transport-time rewrites.
4331 33. The info_callback() function passed to OpenSSL is set up with type void
4332 (*)(SSL *, int, int), as described somewhere. However, when calling the
4333 function (actually a macro) that sets it up, the type void(*)() is
4334 expected. I've put in a cast to prevent warnings from picky compilers.
4336 34. If a DNS black list lookup found a CNAME record, but there were no A
4337 records associated with the domain it pointed at, Exim crashed.
4339 35. If a DNS black list lookup returned more than one A record, Exim ignored
4340 all but the first. It now scans all returned addresses if a particular IP
4341 value is being sought. In this situation, the contents of the
4342 $dnslist_value variable are a list of all the addresses, separated by a
4345 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4346 to accept a host name and all its aliases if the forward lookup for any of
4347 them yielded the IP address of the incoming connection. Now it accepts only
4348 those names whose forward lookup yields the correct IP address. Any other
4349 names are discarded. This closes a loophole whereby a rogue DNS
4350 administrator could create reverse DNS records to break through a
4351 wildcarded host restriction in an ACL.
4353 37. If a user filter or a system filter that ran in a subprocess used any of
4354 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4355 the wrong values were passed to the pipe command ($thisaddress had the
4356 value of $0, $0 had the value of $1, etc). This bug was introduced by
4357 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4359 38. Improved the line breaking for long SMTP error messages from ACLs.
4360 Previously, if there was no break point between 40 and 75 characters, Exim
4361 left the rest of the message alone. Two changes have been made: (a) I've
4362 reduced the minimum length to 35 characters; (b) if it can't find a break
4363 point between 35 and 75 characters, it looks ahead and uses the first one
4364 that it finds. This may give the occasional overlong line, but at least the
4365 remaining text gets split now.
4367 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4368 file descriptors might be low, and that setting 1000 would always raise it.
4369 It turns out that in some environments, the limit is already over 1000 and
4370 that lowering it causes trouble. So now Exim takes care not to decrease it.
4372 40. When delivering a message, the value of $return_path is set to $sender_
4373 address at the start of routing (routers may change the value). By an
4374 oversight, this default was not being set up when an address was tested by
4375 -bt or -bv, which affected the outcome if any router or filter referred to
4378 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4379 log only when "message" or "log_message" are set. However, if one of the
4380 conditions was an address verification, or a call to a nested ACL, the
4381 messages generated by the underlying test were being passed through. This
4382 no longer happens. The underlying message is available in $acl_verify_
4383 message for both "message" and "log_message" expansions, so it can be
4384 passed through if needed.
4386 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4387 new expansion $bh_ to give the encoded byte string without charset
4388 translation. Translation happens only if iconv() is available; HAVE_ICONV
4389 indicates this at build time. HEADERS_CHARSET gives the charset to
4390 translate to; headers_charset can change it in the configuration, and
4391 "headers charset" can change it in an individual filter file.
4393 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4394 that creates RFC 2047 encoded "words" labels them as that charset instead
4395 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4396 expansion operator; (ii) when Exim creates a From: line for a local
4397 message; (iii) when a header line is rewritten to include a "phrase" part.
4399 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4400 buggy, causing it to skip the first lines of messages whose message ID
4401 ended in 'D'. This would not have bitten before Exim release 4.14, because
4402 message IDs were unlikely to end in 'D' before then. The effect was to have
4403 incorrect size information for certain domains.
4405 45. #include "config.h" was missing at the start of the crypt16.c module. This
4406 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4409 46. If there was a timeout during a "random" callout check, Exim treated it as
4410 a failure of the random address, and carried on sending RSET and the real
4411 address. If the delay was just some slowness somewhere, the response to the
4412 original RCPT would be taken as a response to RSET and so on, causing
4413 mayhem of various kinds.
4415 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4416 when I implemented it. It didn't allow for the fact that some option values
4417 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4418 the right test for positive values.
4420 48. Domain names in DNS records are case-independent. Exim always looks them up
4421 in lower case. Some resolvers return domain names in exactly the case they
4422 appear in the zone file, that is, they may contain uppercase letters. Not
4423 all resolvers do this - some return always lower case. Exim was treating a
4424 change of case by a resolver as a change of domain, similar to a widening
4425 of a domain abbreviation. This triggered its re-routing code and so it was
4426 trying to route what was effectively the same domain again. This normally
4427 caused routing to fail (because the router wouldn't handle the domain
4428 twice). Now Exim checks for this case specially, and just changes the
4429 casing of the domain that it ultimately uses when it transmits the message
4432 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4435 50. If a filter generated a file delivery with a non-absolute name (possible if
4436 no home directory exists for the router), the forbid_file option was not
4439 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4440 the existing equality matching.
4442 52. Exim was using ints instead of ino_t variables in some places where it was
4443 dealing with inode numbers.
4445 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4446 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4447 called TMPDIR, and if it finds it is different, it changes its value.
4449 54. The smtp_printf() function is now made available to local_scan() so
4450 additional output lines can be written before returning. There is also an
4451 smtp_fflush() function to enable the detection of a dropped connection.
4452 The variables smtp_input and smtp_batched_input are exported to
4455 55. Changed the default runtime configuration: the message "Unknown user"
4456 has been removed from the ACL, and instead placed on the localuser router,
4457 using the cannot_route_message feature. This means that any verification
4458 failures that generate their own messages won't get overridden. Similarly,
4459 the "Unrouteable address" message that was in the ACL for unverifiable
4460 relay addresses has also been removed.
4462 56. Added hosts_avoid_esmtp to the smtp transport.
4464 57. The exicyclog script was not checking for the esoteric option
4465 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4466 will work only if exicyclog is run under the appropriate euid.
4468 58. Following a discussion on the list, the rules by which Exim recognises line
4469 endings on incoming messages have been changed. The -dropcr and drop_cr
4470 options are now no-ops, retained only for backwards compatibility. The
4471 following line terminators are recognized: LF CRLF CR. However, special
4472 processing applies to CR:
4474 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4475 nor a local message in the state where . is a terminator.
4477 (ii) If a bare CR is encountered in a header line, an extra space is added
4478 after the line terminator so as not to end the header. The reasoning
4479 behind this is that bare CRs in header lines are most likely either
4480 to be mistakes, or people trying to play silly games.
4482 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4483 was being incorrectly given as 18 bytes larger than it should have been.
4484 This is a VOB (very old bug).
4486 60. This may never have affected anything current, but just in case it has:
4487 When the local host is found other than at the start of a list of hosts,
4488 the local host, those with the same MX, and any that follow, are discarded.
4489 When the list in question was part of a longer list of hosts, the following
4490 hosts (not currently being processed) were also being discarded. This no
4491 longer happens. I'm not sure if this situation could ever has previously
4494 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4497 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4498 header. This is something that is recommended in a new Internet Draft, and
4499 is something that is documented as being done by Sendmail. There are two
4500 possible values. For messages generated by the autoreply transport, Exim
4503 Auto-Submitted: auto-replied
4505 whereas for all other generated messages (e.g. bounces) it adds
4507 Auto-Submitted: auto-generated
4509 63. The "personal" condition in filters now includes a test for the
4510 Auto-Submitted: header. If it contains the string "auto-" the message it
4511 not considered personal.
4513 64. Added rcpt_include_affixes as a generic transport option.
4515 65. Added queue_only_override (default true).
4517 66. Added the syslog_duplication option.
4519 67. If what should have been the first header line of a message consisted of
4520 a space followed by a colon, Exim was mis-interpreting it as a header line.
4521 It isn't of course - it is syntactically invalid and should therefore be
4522 treated as the start of the message body. The misbehaviour could have
4523 caused a number of strange effects, including loss of data in subsequent
4524 header lines, and spool format errors.
4526 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4527 client host had authenticated. This control can now be exercised by an ACL
4528 for more flexibility.
4530 69. By default, callouts do not happen when testing with -bh. There is now a
4531 variant, -bhc, which does actually run the callout code, including
4532 consulting and updating the callout cache.
4534 70. Added support for saslauthd authentication, courtesy of Alexander
4537 71. If statvfs() failed on the spool or log directories while checking their
4538 size for availability, Exim confusingly gave the error "space shortage".
4539 Furthermore, in debugging mode it crashed with a floating point exception.
4540 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4541 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4542 serious problem, Exim now writes to the main and panic logs when this
4543 happens, with details of the failure. It then refuses to accept the
4544 incoming message, giving the message "spool directory problem" or "log
4545 directory problem" with a 421 code for SMTP messages.
4547 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4548 0, 1, and 2 exist, because some OS complain for execs without them (see
4549 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4550 descriptors. However, the code omitted to check that the open succeeded,
4551 causing mysterious errors if for some reason the permissions on /dev/null
4552 got screwed. Now Exim writes a message to the main and panic logs, and
4553 bombs out if it can't open /dev/null.
4555 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4556 interact so that it is all more flexible. It is supposed to remain
4557 backwards compatible. Also added extra_local_interfaces.
4559 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4560 to bomb out with an assertion failure - to the client this appears as a
4561 connection drop. This problem occurs in the part of the code that was taken
4562 from the Samba project. Fortunately, the assertion is in a very simple
4563 function, so I have fixed this by reproducing the function inline in the
4564 one place where it is called, and arranging for authentication to fail
4565 instead of killing the process with assert().
4567 75. The SPA client code was not working when the server requested OEM rather
4568 than Unicode encoding.
4570 76. Added code to make require_files with a specific uid setting more usable in
4571 the case where statting the file as root fails - usually a non-root-mounted
4572 NFS file system. When this happens and the failure is EACCES, Exim now
4573 forks a subprocess and does the per-uid checking as the relevant uid.
4575 77. Added process_log_path.
4577 78. If log_file_path was not explicitly set, a setting of check_log_space or
4578 check_log_inodes was ignored.
4580 79. If a space check for the spool or log partitions fails, the incident is now
4581 logged. Of course, in the latter case the data may get lost...
4583 80. Added the %p formatting code to string_format() so that it can be used to
4584 print addresses in debug_print(). Adjusted all the address printing in the
4585 debugging in store.c to use %p rather than %d.
4587 81. There was a concern that a line of code in smtp_in.c could overflow a
4588 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4589 initially expressed, the concern was not well-founded, because trailing
4590 spaces are removed early. However, if the trailing spaces were followed by
4591 a NULL, they did not get removed, so the overflow was possible. Two fixes
4594 (a) I re-wrote the offending code in a cleaner fashion.
4595 (b) If an incoming SMTP command contains a NULL character, it is rejected
4598 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4599 runs initgroups(), so that if the Exim user is in any additional groups,
4600 they will be used during message reception.
4606 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob