1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.474 2007/02/07 11:24:56 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.
90 PH/21 Long custom messages for fakedefer and fakereject are now split up
91 into multiline reponses in the same way that messages for "deny" and
92 other ACL rejections are.
98 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
99 fixed by 4.65/MH/01 (is this a record?) are fixed:
101 (i) An empty string was always treated as zero by the numeric comparison
102 operators. This behaviour has been restored.
104 (ii) It is documented that the numeric comparison operators always treat
105 their arguments as decimal numbers. This was broken in that numbers
106 starting with 0 were being interpreted as octal.
108 While fixing these problems I realized that there was another issue that
109 hadn't been noticed. Values of message_size_limit (both the global option
110 and the transport option) were treated as octal if they started with 0.
111 The documentation was vague. These values are now always treated as
112 decimal, and I will make that clear in the documentation.
118 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
119 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
122 MH/01 Don't check that the operands of numeric comparison operators are
123 integers when their expansion is in "skipping" mode (fixes bug
124 introduced by 4.64-PH/07).
126 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
127 child addresses, Exim now panics and dies. Previously, because the count
128 is held in a short int, deliveries were likely to be lost. As such a
129 large number of recipients for a single message is ridiculous
130 (performance will be very, very poor), I have chosen to impose a limit
131 rather than extend the field.
137 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
138 leftover -K file (the existence of which was triggered by #402).
139 While we were at it, introduced process PID as part of the -K
140 filename. This should rule out race conditions when creating
143 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
144 processing considerably. Previous code took too long for large mails,
145 triggering a timeout which in turn triggers #401.
147 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
148 in the DK code in transports.c. sendfile() is not really portable,
149 hence the _LINUX specificness.
151 TF/01 In the add_headers option to the mail command in an Exim filter,
152 there was a bug that Exim would claim a syntax error in any
153 header after the first one which had an odd number of characters
156 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
157 callout verification, Exim cached a "reject" for the entire domain. This
158 is correct for most verifications, but it is not correct for a recipient
159 verification with use_sender or use_postmaster set, because in that case
160 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
161 case of MAIL FROM:<> rejection from other early rejections (e.g.
162 rejection of HELO). When verifying a recipient using a non-null MAIL
163 address, the cache is ignored if it shows MAIL FROM:<> rejection.
164 Whatever the result of the callout, the value of the domain cache is
165 left unchanged (for any other kind of callout, getting as far as trying
166 RCPT means that the domain itself is ok).
168 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
171 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
172 manifest itself as EPIPE rather than ECONNECT. When tidying away a
173 session, the daemon ignores ECONNECT errors and logs others; it now
174 ignores EPIPE as well.
176 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
177 (quoted-printable decoding).
179 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
180 later the small subsequent patch to fix an introduced bug.
182 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
184 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
186 PH/08 An error is now given if message_size_limit is specified negative.
188 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
189 to be given (somewhat) arbitrary names.
191 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
194 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
195 miscellaneous code fixes
197 PH/10 Added the log_reject_target ACL modifier to specify where to log
200 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
201 hostname. This is wrong, because it relates to the incoming message (and
202 probably the interface on which it is arriving) and not to the outgoing
203 callout (which could be using a different interface). This has been
204 changed to use the value of the helo_data option from the smtp transport
205 instead - this is what is used when a message is actually being sent. If
206 there is no remote transport (possible with a router that sets up host
207 addresses), $smtp_active_hostname is used.
209 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
210 tweaks were necessary in order to get it to work (see also 21 below):
211 (a) The code assumed that strncpy() returns a negative number on buffer
212 overflow, which isn't the case. Replaced with Exim's string_format()
214 (b) There were several signed/unsigned issues. I just did the minimum
215 hacking in of casts. There is scope for a larger refactoring.
216 (c) The code used strcasecmp() which is not a standard C function.
217 Replaced with Exim's strcmpic() function.
218 (d) The code set only $1; it now sets $auth1 as well.
219 (e) A simple test gave the error "authentication client didn't specify
220 service in request". It would seem that Dovecot has changed its
221 interface. Fortunately there's a specification; I followed it and
222 changed what the client sends and it appears to be working now.
224 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
227 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
228 address A is aliased to B and C, where B exists and C does not. Without
229 -v the output is "A verified" because verification stops after a
230 successful redirection if more than one address is generated. However,
231 with -v the child addresses are also verified. Exim was outputting "A
232 failed to verify" and then showing the successful verification for C,
233 with its parentage. It now outputs "B failed to verify", showing B's
234 parentage before showing the successful verification of C.
236 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
237 look up a TXT record in a specific list after matching in a combined
240 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
241 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
242 they consult the DNS. I had assumed they would set it the way they
243 wanted; and indeed my experiments on Linux seem to show that in some
244 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
245 To be on the safe side, however, I have now made the interface to
246 host_find_byname() similar to host_find_bydns(), with an argument
247 containing the DNS resolver options. The host_find_byname() function now
248 sets these options at its start, just as host_find_bydns() does. The smtp
249 transport options dns_qualify_single and dns_search_parents are passed to
250 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
251 of host_find_byname() use the default settings of RES_DEFNAMES
252 (qualify_single) but not RES_DNSRCH (search_parents).
254 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
255 spool_read_header() do less string testing, by means of a preliminary
256 switch on the second character of optional "-foo" lines. (This is
257 overdue, caused by the large number of possibilities that now exist.
258 Originally there were few.) While I was there, I also converted the
259 str(n)cmp tests so they don't re-test the leading "-" and the first
260 character, in the hope this might squeeze out yet more improvement.
262 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
263 flag allowing group syntax was set by the header_syntax check but not
264 turned off, possible causing trouble later; (2) The flag was not being
265 set at all for the header_verify test, causing "group"-style headers to
266 be rejected. I have now set it in this case, and also caused header_
267 verify to ignore an empty address taken from a group. While doing this, I
268 came across some other cases where the code for allowing group syntax
269 while scanning a header line wasn't quite right (mostly, not resetting
270 the flag correctly in the right place). These bugs could have caused
271 trouble for malformed header lines. I hope it is now all correct.
273 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
274 with the "reply" argument non-NULL. The code, however (which originally
275 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
276 but it didn't always do it. This confused somebody who was copying the
277 code for some other use. I have removed all the tests.
279 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
280 feature that was used to support insecure browsers during the U.S. crypto
281 embargo. It requires special client support, and Exim is probably the
282 only MTA that supported it -- and would never use it because real RSA is
283 always available. This code has been removed, because it had the bad
284 effect of slowing Exim down by computing (never used) parameters for the
285 RSA_EXPORT functionality.
287 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
288 authenticator to fail if there's a tab character in the incoming data
289 (there should never be unless someone is messing about, as it's supposed
290 to be base64-encoded). Also added, on Timo's advice, the "secured" option
291 if the connection is using TLS or if the remote IP is the same as the
292 local IP, and the "valid-client-cert option" if a client certificate has
295 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
296 authenticators. This can be used for authorization after authentication
297 succeeds. (In the case of plaintext, it servers for both authentication
300 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
301 if any retry times were supplied.
303 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
304 connection, where there is no client IP address to check. In this
305 situation, the verify now always succeeds.
307 PH/25 Applied John Jetmore's -Mset patch.
309 PH/26 Added -bem to be like -Mset, but loading a message from a file.
311 PH/27 In a string expansion for a processed (not raw) header when multiple
312 headers of the same name were present, leading whitespace was being
313 removed from all of them, but trailing whitespace was being removed only
314 from the last one. Now trailing whitespace is removed from each header
315 before concatenation. Completely empty headers in a concatenation (as
318 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
319 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
321 PH/29 [Removed. This was a change that I later backed out, and forgot to
322 correct the ChangeLog entry (that I had efficiently created) before
323 committing the later change.]
325 PH/30 Exim was sometimes attempting to deliver messages that had suffered
326 address errors (4xx response to RCPT) over the same connection as other
327 messages routed to the same hosts. Such deliveries are always "forced",
328 so retry times are not inspected. This resulted in far too many retries
329 for the affected addresses. The effect occurred only when there were more
330 hosts than the hosts_max_try setting in the smtp transport when it had
331 the 4xx errors. Those hosts that it had tried were not added to the list
332 of hosts for which the message was waiting, so if all were tried, there
333 was no problem. Two fixes have been applied:
335 (i) If there are any address or message errors in an SMTP delivery, none
336 of the hosts (tried or untried) are now added to the list of hosts
337 for which the message is waiting, so the message should not be a
338 candidate for sending over the same connection that was used for a
339 successful delivery of some other message. This seems entirely
340 reasonable: after all the message is NOT "waiting for some host".
341 This is so "obvious" that I'm not sure why it wasn't done
342 previously. Hope I haven't missed anything, but it can't do any
343 harm, as the worst effect is to miss an optimization.
345 (ii) If, despite (i), such a delivery is accidentally attempted, the
346 routing retry time is respected, so at least it doesn't keep
347 hammering the server.
349 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
350 in ${readsocket because some servers need this prod.
352 PH/32 Added some extra debug output when updating a wait-xxx database.
354 PH/33 The hint "could be header name not terminated by colon", which has been
355 given for certain expansion errors for a long time, was not being given
356 for the ${if def:h_colon_omitted{... case.
358 PH/34 The spec says: "With one important exception, whenever a domain list is
359 being scanned, $domain contains the subject domain." There was at least
360 one case where this was not true.
362 PH/35 The error "getsockname() failed: connection reset by peer" was being
363 written to the panic log as well as the main log, but it isn't really
364 panic-worthy as it just means the connection died rather early on. I have
365 removed the panic log writing for the ECONNRESET error when getsockname()
368 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
369 runs only) independently of the message's sender address. This meant
370 that, if the 4xx error was in fact related to the sender, a different
371 message to the same recipient with a different sender could confuse
372 things. In particualar, this can happen when sending to a greylisting
373 server, but other circumstances could also provoke similar problems.
374 I have changed the default so that the retry time for these errors is now
375 based a combination of the sender and recipient addresses. This change
376 can be overridden by setting address_retry_include_sender=false in the
379 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
380 remote server are returned as part of bounce messages. This was not
381 happening for LMTP over a pipe (the lmtp transport), but now it is the
382 same for both kinds of LMTP.
384 PH/38 Despite being documented as not happening, Exim was rewriting addresses
385 in header lines that were in fact CNAMEs. This is no longer the case.
387 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
388 and queue runs started by the daemon processed all messages. This has
389 been fixed so that -R and -S can now usefully be given with -q<time>.
391 PH/40 Import PCRE release 6.7 (fixes some bugs).
393 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
395 PH/42 Give an error if -q is specified more than once.
397 PH/43 Renamed the variables $interface_address and $interface_port as
398 $received_ip_address and $received_port, to make it clear that these
399 values apply to message reception, and not to the outgoing interface when
400 a message is delivered. (The old names remain recognized, of course.)
402 PH/44 There was no timeout on the connect() call when using a Unix domain
403 socket in the ${readsocket expansion. There now is.
405 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
406 be meaningful with "accept".
408 SC/01 Eximstats V1.43
409 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
411 SC/02 Eximstats V1.44
412 Use a glob alias rather than an array ref in the generated
413 parser. This improves both readability and performance.
415 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
416 Collect SpamAssassin and rejection statistics.
417 Don't display local sender or destination tables unless
418 there is data to show.
419 Added average volumes into the top table text output.
421 SC/04 Eximstats V1.46
422 Collect data on the number of addresses (recipients)
423 as well as the number of messages.
425 SC/05 Eximstats V1.47
426 Added 'Message too big' to the list of mail rejection
427 reasons (thanks to Marco Gaiarin).
429 SC/06 Eximstats V1.48
430 Mainlog lines which have GMT offsets and are too short to
431 have a flag are now skipped.
433 SC/07 Eximstats V1.49 (Alain Williams)
434 Added the -emptyok flag.
436 SC/08 Eximstats V1.50
437 Fixes for obtaining the IP address from reject messages.
439 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
440 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
441 whitesspace changes from 4.64-PH/27
443 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
446 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
447 are found, allow negative numbers in numeric criteria)
449 JJ/06 exipick.20061117.2, added new $message_body_missing variable
451 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
452 to match changes made in 4.64-PH/43
454 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
456 PH/47 Put in an explicit test for a DNS lookup of an address record where the
457 "domain" is actually an IP address, and force a failure. This locks out
458 those revolvers/nameservers that support "A-for-A" lookups, in
459 contravention of the specifications.
461 PH/48 When a host name was looked up from an IP address, and the subsequent
462 forward lookup of the name timed out, the host name was left in
463 $sender_host_name, contrary to the specification.
465 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
466 restricted to single-key lookups, Exim was not diagnosing an error if
467 * or *@ was used with a query-style lookup.
469 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
471 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
472 long ago, but noone interested enough thought of it. Let's just say that
473 the "1.1" means that there are some new functions that weren't there at
474 some point in the past.
476 PH/51 Error processing for expansion failure of helo_data from an smtp
477 transport during callout processing was broken.
479 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
480 tested/used via the -bh/-bhc/-bs options.
482 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
483 bug, fixed in subsequent PCRE releases).
485 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
486 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
488 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
494 SC/01 Use a glob alias rather than an array ref in eximstats generated
495 parser. This improves both readability and performance.
497 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
498 Don't display local sender or destination tables in eximstats unless
499 there is data to show.
500 Added average volumes into the eximstats top table text output.
502 SC/03 Collect data on the number of addresses (recipients) as well
503 as the number of messages in eximstats.
505 TF/01 Correct an error in the documentation for the redirect router. Exim
506 does (usually) call initgroups() when daemonizing.
508 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
509 with consistent privilege compared to when running as a daemon.
511 TF/03 Note in the spec that $authenticated_id is not set for local
512 submissions from trusted users.
514 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
515 Thanks to Dean Brooks <dean@iglou.com> for the patch.
517 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
518 by adding some example configuration directives to the default
519 configuration file. A little bit of work is required to uncomment the
520 directives and define how usernames and passwords are checked, but
521 there is now a framework to start from.
523 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
524 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
525 without this. I don't know how relevant this is to other LDAP libraries.
527 PH/02 Add the verb name to the "unknown ACL verb" error.
529 PH/03 Magnus Holmgren's patch for filter_prepend_home.
531 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
533 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
534 directory not expanded when it should be if an expanded home directory
535 was set for the address (which is overridden by the transport).
537 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
540 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
541 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
542 because it is too late at that time, and has no effect.
544 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
545 security issue with \' (bugzilla #107). I could not use the
546 PQescapeStringConn() function, because it needs a PGconn value as one of
549 PH/08 When testing addresses using -bt, indicate those final addresses that
550 are duplicates that would not cause an additional delivery. At least one
551 person was confused, thinking that -bt output corresponded to deliveries.
552 (Suppressing duplicates isn't a good idea as you lose the information
553 about possibly different redirections that led to the duplicates.)
555 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
556 systems where poll() doesn't work, in particular OS X.
558 PH/10 Added more information to debugging output for retry time not reached.
560 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
561 operations in malware.c.
563 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
566 PH/13 If write_rejectlog was set false when logging was sent to syslog with
567 syslog_duplication set false, log lines that would normally be written
568 both the the main log and to the reject log were not written to syslog at
571 PH/14 In the default configuration, change the use of "message" in ACL warn
572 statements to "add_header".
574 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
575 not followed by a command (e.g. "seen endif").
577 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
578 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
581 PH/17 Added extra conditions to the default value of delay_warning_condition
585 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
586 { match{$h_precedence:}{(?i)bulk|list|junk} } \
587 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
590 The Auto-Submitted: and various List- headers are standardised, whereas I
591 don't think Precedence: ever was.
593 PH/18 Refactored debugging code in route_finduser() to show more information,
594 in particular, the error code if getpwnam() issues one.
596 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
597 This is apparently needed in addition to the PH/07 change above to avoid
598 any possible encoding problems.
600 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
601 but not after initializing Perl.
603 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
604 output them only if debugging. By default they are written stderr,
605 apparently, which is not desirable.
607 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
610 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
613 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
615 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
616 authenticated or an ident call has been made. Suppress the default
617 values for $authenticated_id and $authenticated_sender (but permit -oMai
618 and -oMas) when testing with -bh.
620 PH/24 Re-jigged the order of the tests in the default configuration so that the
621 tests for valid domains and recipients precede the DNS black list and CSA
622 tests, on the grounds that those ones are more expensive.
624 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
625 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
626 This bug exists in every version of Exim that I still have, right back to
629 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
630 However, an attempt to turn on case-sensitivity in a regex key by
631 including (?-i) didn't work because the subject string was already
632 lowercased, and the effects were non-intuitive. It turns out that a
633 one-line patch can be used to allow (?-i) to work as expected.
639 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
640 other effects) broke the use of negated acl sub-conditions.
642 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
645 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
646 "Deny" causes Exim to reject the incoming connection with a 554 error.
647 Unfortunately, if there is a major crisis, such as a disk failure,
648 tcp-wrappers gives "deny", whereas what one would like would be some
649 kind of temporary error. A kludge has been added to help with this.
650 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
651 554 error is used if errno is still zero or contains ENOENT (which occurs
652 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
655 PH/03 Add -lutil to the default FreeBSD LIBS setting.
657 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
658 errors. Otherwise a message that provokes a temporary error (when other
659 messages do not) can cause a whole host to time out.
661 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
662 addresses were routed directly to files or pipes from a redirect router.
663 File deliveries just didn't batch; pipe deliveries might have suffered
666 PH/06 A failure to get a lock for a hints database would erroneously always say
667 "Failed to get write lock", even when it was really a read lock.
669 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
670 of 0600. This has been changed to use the value of the lockfile_mode
671 option (which defaults to 0600).
673 PH/08 Applied small patch from the Sieve maintainer.
675 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
676 folder from quota calculations, a direct delivery into this folder messed
677 up the contents of the maildirsize file. This was because the regex was
678 used only to exclude .Trash (or whatever) when the size of the mailbox
679 was calculated. There was no check that a delivery was happening into an
680 excluded directory. This bug has been fixed by ignoring all quota
681 processing for deliveries into excluded directories.
683 PH/10 Added the maildirfolder_create_regex option to appendfile.
689 PH/01 The code for finding all the local interface addresses on a FreeBSD
690 system running IPv6 was broken. This may well have applied to all BSD
691 systems, as well as to others that have similar system calls. The broken
692 code found IPv4 interfaces correctly, but gave incorrect values for the
693 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
694 that it would not match correctly against @[] and not recognize the IPv6
697 PH/02 The ipliteral router was not recognizing addresses of the form user@
698 [ipv6:....] because it didn't know about the "ipv6:" prefix.
700 PH/03 Added disable_ipv6.
702 PH/04 Changed $reply_address to use the raw form of the headers instead of the
703 decoded form, because it is most often used to construct To: headers
704 lines in autoreplies, and the decoded form may well be syntactically
705 invalid. However, $reply_address has leading white space removed, and all
706 newlines turned into spaces so that the autoreply transport does not
709 PH/05 If group was specified without a user on a router, and no group or user
710 was specified on a transport, the group from the router was ignored.
712 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
713 for visible compile-time settings that can be used to change these
714 numbers, for those that want even more. Backwards compatibility with old
715 spool files has been maintained. However, going back to a previous Exim
716 release will lost any variables that are in spool files.
718 PH/07 Two small changes when running in the test harness: increase delay when
719 passing a TCP/IP connection to a new process, in case the original
720 process has to generate a bounce, and remove special handling of
721 127.0.0.2 (sic), which is no longer necessary.
723 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
724 be the same on different OS.
726 PH/09 Moved a debug statement in filter processing to avoid a race problem when
729 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
730 whether --show-vars was specified or not
732 JJ/02 exipick: Added support for new ACL variable spool format introduced
735 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
736 syntactically invalid From: or Reply-to: line, and a filter used this to
737 generate an autoreply, and therefore failed to obtain an address for the
738 autoreply, Exim could try to deliver to a non-existent relative file
739 name, causing unrelated and misleading errors. What now happens is that
740 it logs this as a hard delivery error, but does not attempt to create a
743 PH/11 The exinext utility has a -C option for testing purposes, but although
744 the given file was scanned by exinext itself; it wasn't being passed on
745 when Exim was called.
747 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
748 an end-of-file indication when reading a command response.
750 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
751 compiled. In many other places in Exim, IPv6 addresses are always
752 recognized, so I have changed this. It also means that IPv4 domain
753 literals of the form [IPV4:n.n.n.n] are now always recognized.
755 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
756 used if the router is not running as root, for example, when verifying at
757 ACL time, or when using -bh. The debugging output from this situation was
758 non-existent - all you got was a failure to exec. I have made two
761 (a) Failures to set uid/gid, the current directory, or a process leader
762 in a subprocess such as that created by queryprogram now generate
763 suitable debugging ouput when -d is set.
765 (b) The queryprogram router detects when it is not running as root,
766 outputs suitable debugging information if -d is set, and then runs
767 the subprocess without attempting to change uid/gid.
769 PH/15 Minor change to Makefile for building test_host (undocumented testing
772 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
773 additional section of a DNS packet that returns MX or SRV records.
774 Instead, it always explicitly searches for A/AAAA records. This avoids
775 major problems that occur when a DNS server includes only records of one
776 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
777 fixed another bug: if SRV records were looked up and the corresponding
778 address records were *not* found in the additional section, the port
779 values from the SRV records were lost.
781 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
782 using the correct key (the original address) when searching the retry
783 rules in order to find which one to use for generating the retry hint.
785 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
786 adding the default one. Similarly, if it contains a Reply-To: header, the
787 errors_reply_to option, if set, is not used.
789 PH/19 When calculating a retry time, Exim used to measure the "time since
790 failure" by looking at the "first failed" field in the retry record. Now
791 it does not use this if it is later than than the arrival time of the
792 message. Instead it uses the arrival time. This makes for better
793 behaviour in cases where some deliveries succeed, thus re-setting the
794 "first failed" field. An example is a quota failure for a huge message
795 when small messages continue to be delivered. Without this change, the
796 "time since failure" will always be short, possible causing more frequent
797 delivery attempts for the huge message than are intended.
798 [Note: This change was subsequently modified - see PH/04 for 4.62.]
800 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
801 $1, $2, $3) because the numerical variables can be reset during some
802 expansion items (e.g. "match"), thereby losing the authentication data.
804 PH/21 Make -bV show the size of off_t variables so that the test suite can
805 decide whether to run tests for quotas > 2G.
807 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
808 mailbox_size, and mailbox_filecount in the appendfile transport. If a
809 filecount value is greater than 2G or if a quota value is greater than 2G
810 on a system where the size of off_t is not greater than 4, a panic error
813 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
814 never match. The debug and -bh output now contains an explicit error
815 message indicating a malformed IPv4 address or mask.
817 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
818 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
821 PH/25 Do not write to syslog when running in the test harness. The only
822 occasion when this arises is a failure to open the main or panic logs
823 (for which there is an explicit test).
825 PH/26 Added the /no_tell option to "control=freeze".
827 PH/27 If a host name lookup failed very early in a connection, for example, if
828 the IP address matched host_lookup and the reverse lookup yielded a name
829 that did not have a forward lookup, an error message of the form "no IP
830 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
831 could be logged. Now it outputs the IP address instead of "NULL".
833 PH/28 An enabling patch from MH: add new function child_open_exim2() which
834 allows the sender and the authenticated sender to be set when
835 submitting a message from within Exim. Since child_open_exim() is
836 documented for local_scan(), the new function should be too.
838 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
839 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
840 results in an empty string is now treated as unset.
842 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
844 PH/31 Added sender_verify_fail logging option.
846 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
847 needed by RFC 822 but not by RFC 2822 was commented out. I have now
848 tidied the source and removed it altogether.
850 PH/33 When a queue run was abandoned because the load average was too high, a
851 log line was always written; now it is written only if the queue_run log
852 selector is set. In addition, the log line for abandonment now contains
853 information about the queue run such as the pid. This is always present
854 in "start" and "stop" lines but was omitted from the "abandon" line.
856 PH/34 Omit spaces between a header name and the colon in the error message that
857 is given when verify = headers_syntax fails (if there are lots of them,
858 the message gets confusing).
860 PH/35 Change the default for dns_check_names_pattern to allow slashes within
861 names, as there are now some PTR records that contain slashes. This check
862 is only to protect against broken name servers that fall over on strange
863 characters, so the fact that it applies to all lookups doesn't matter.
865 PH/36 Now that the new test suite is complete, we can remove some of the
866 special code in Exim that was needed for the old test suite. For example,
867 sorting DNS records because real resolvers return them in an arbitrary
868 order. The new test suite's fake resolver always returns records in the
871 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
872 bounces) except when queue_only is set, to avoid logging races between
873 the different processes.
875 PH/38 Panic-die if .include specifies a non-absolute path.
877 PH/39 A tweak to the "H" retry rule from its user.
879 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
880 a label. They prevented compilation on older perls.
882 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
883 a warning to be raised on newish perls.
885 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
886 on queue. Changes to match documented behaviour of showing count of
887 messages matching specified criteria.
889 PH/40 Changed the default ident timeout from 30s to 5s.
891 PH/41 Added support for the use of login_cap features, on those BSD systems
892 that have them, for controlling the resources used by pipe deliveries.
894 PH/42 The content-scanning code uses fopen() to create files in which to put
895 message data. Previously it was not paying any attention to the mode of
896 the files. Exim runs with umask(0) because the rest of the code creates
897 files with open(), and sets the required mode explicitly. Thus, these
898 files were ending up world-writeable. This was not a big issue, because,
899 being within the spool directory, they were not world-accessible. I have
900 created a function called modefopen, which takes an additional mode
901 argument. It sets umask(777), creates the file, chmods it to the required
902 mode, then resets the umask. All the relevant calls to fopen() in the
903 content scanning code have been changed to use this function.
905 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
906 to 24 hours. This avoids potential overflow problems when processing G
907 and H retry rules. I suspect nobody ever tinkers with this value.
909 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
911 PH/45 When the plaintext authenticator is running as a client, the server's
912 challenges are checked to ensure they are valid base64 strings. By
913 default, the authentication attempt is cancelled if an invalid string is
914 received. Setting client_ignore_invalid_base64 true ignores these errors.
915 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
916 they are received. Thus, the responses can be made to depend on the
917 challenges. If an invalid string is ignored, an empty string is placed in
920 PH/46 Messages that are created by the autoreply transport now contains a
921 References: header, in accordance with RFCs 2822 and 3834.
923 PH/47 Added authenticated_sender_force to the smtp transport.
925 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
927 PH/49 Installed latest patch from the Sieve maintainer.
929 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
930 was also set, the appendfile transport was unnecessarily scanning a
931 directory of message files (e.g. for maildir delivery) to find the count
932 of files (along with the size), even though it did not need this
933 information. It now does the scan only if it needs to find either the
934 size of the count of files.
936 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
938 PH/52 Two bugs concerned with error handling when the smtp transport is
941 (i) Exim was not creating retry information for temporary errors given
942 for individual recipients after the DATA command when the smtp transport
943 was used in LMTP mode. This meant that they could be retried too
944 frequently, and not timed out correctly.
946 (ii) Exim was setting the flag that allows error details to be returned
947 for LMTP errors on RCPT commands, but not for LMTP errors for individual
948 recipients that were returned after the DATA command.
950 PH/53 This is related to PH/52, but is more general: for any failing address,
951 when detailed error information was permitted to be returned to the
952 sender, but the error was temporary, then after the final timeout, only
953 "retry timeout exceeded" was returned. Now it returns the full error as
954 well as "retry timeout exceeded".
956 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
957 do this, and (what is worse) MTAs that accept it.
959 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
960 will now be deprecated.
962 PH/56 New os.c-cygwin from the Cygwin maintainer.
964 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
965 formats (previously only available in exim formats via -bpr, -bpru,
966 and -bpra. Now also available in native and exiqgrep formats)
968 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
969 with very large, slow to parse queues
971 JJ/08 exipick: added ! as generic prefix to negate any criteria format
973 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
975 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
976 responses to authentication challenges, though it was showing the
977 challenges; (ii) I've removed the CR characters from the debug output for
980 PH/58 Allow for the insertion of a newline as well as a space when a string
981 is turned into more than one encoded-word during RFC 2047 encoding. The
982 Sieve code now uses this.
984 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
985 data_4xx, lost_connection, tls_required.
987 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
988 message was being sent as an SMTP response.
990 PH/61 Add -l and -k options to exicyclog.
992 PH/62 When verifying, if an address was redirected to one new address, so that
993 verification continued, and the new address failed or deferred after
994 having set something in $address_data, the value of $address_data was not
995 passed back to the ACL. This was different to the case when no
996 redirection occurred. The value is now passed back in both cases.
998 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
999 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1000 use login_cap.h, so on its own it isn't the distinguishing feature. The
1001 new name refers directly to the setclassresources() function.
1003 PH/65 Added configuration files for NetBSD3.
1005 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1007 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1008 is preferred over IPv4.
1010 PH/68 The bounce_return_message and bounce_return_body options were not being
1011 honoured for bounces generated during the reception of non-SMTP messages.
1012 In particular, this applied to messages rejected by the ACL. This bug has
1013 been fixed. However, if bounce_return_message is true and bounce_return_
1014 body is false, the headers that are returned for a non-SMTP message
1015 include only those that have been read before the error was detected.
1016 (In the case of an ACL rejection, they have all been read.)
1018 PH/69 The HTML version of the specification is now built in a directory called
1019 spec_html instead of spec.html, because the latter looks like a path with
1020 a MIME-type, and this confuses some software.
1022 PH/70 Catch two compiler warnings in sieve.c.
1024 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1025 function verify_get_ident() calls ip_connect() to connect a socket, but
1026 if the "connect()" function timed out, ip_connect() used to close the
1027 socket. However, verify_get_ident() also closes the socket later, and in
1028 between Exim writes to the log, which may get opened at this point. When
1029 the socket was closed in ip_connect(), the log could get the same file
1030 descriptor number as the socket. This naturally causes chaos. The fix is
1031 not to close the socket in ip_connect(); the socket should be closed by
1032 the function that creates it. There was only one place in the code where
1033 this was missing, in the iplookup router, which I don't think anybody now
1034 uses, but I've fixed it anyway.
1036 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1037 well as to direct DNS lookups. Otherwise the handling of names in host
1038 lists is inconsistent and therefore confusing.
1044 PH/01 Two changes to the default runtime configuration:
1046 (1) Move the checks for relay_from_hosts and authenticated clients from
1047 after to before the (commented out) DNS black list checks.
1049 (2) Add control=submission to the relay_from_hosts and authenticated
1050 clients checks, on the grounds that messages accepted by these
1051 statements are most likely to be submissions.
1053 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1055 (1) Generate an error if the third argument for the ${prvs expansion is
1058 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1061 (3) Reset the variables that are obtained from the first argument of
1062 ${prvscheck and used in the second argument before leaving the code,
1063 because their memory is reclaimed, so using them afterwards may do
1066 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1067 one (it's much easier than Tom thought :-).
1069 (5) Because of (4), we can now allow for the use of $prvscheck_result
1070 inside the third argument.
1072 PH/03 For some reason, the default setting of PATH when running a command from
1073 a pipe transport was just "/usr/bin". I have changed it to
1076 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1077 anything to be listed in the output from -bV.
1079 PH/05 When a filter generated an autoreply, the entire To: header line was
1080 quoted in the delivery log line, like this:
1082 => >A.N.Other <ano@some.domain> <original@ddress> ...
1084 This has been changed so that it extracts the operative address. There
1085 may be more than one such address. If so, they are comma-separated, like
1088 => >ano@some.domain,ona@other.domain <original@ddress> ...
1090 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1091 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1092 not being looked up in the rDNS to get a host name, Exim was showing the
1093 IP address twice in Received: lines, even though the IP addresses were
1094 identical. For example:
1096 Received: from [1.2.3.4] (helo=[1.2.3.4])
1098 However, if the real host name was known, it was omitting the HELO data
1099 if it matched the actual IP address. This has been tidied up so that it
1100 doesn't show the same IP address twice.
1102 PH/07 When both +timestamp and +memory debugging was on, the value given by
1103 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1104 called by the debug printing, thereby overwriting the timestamp buffer.
1105 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1107 PH/08 When the original message was included in an autoreply transport, it
1108 always said "this is a copy of the message, including all the headers",
1109 even if body_only or headers_only was set. It now gives an appropriate
1112 PH/09 Applied a patch from the Sieve maintainer which:
1114 o fixes some comments
1115 o adds the (disabled) notify extension core
1116 o adds some debug output for the result of if/elsif tests
1117 o points to the current vacation draft in the documentation
1118 and documents the missing references header update
1122 o fixes a bug in processing the envelope test (when testing
1123 multiple envelope elements, the last element determinted the
1126 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1127 Electronic Mail") by including:
1129 Auto-submitted: auto-generated
1131 in the messages that it generates (bounce messages and others, such as
1132 warnings). In the case of bounce messages for non-SMTP mesages, there was
1133 also a typo: it was using "Auto_submitted" (underscore instead of
1134 hyphen). Since every message generated by Exim is necessarily in response
1135 to another message, thes have all been changed to:
1137 Auto-Submitted: auto-replied
1139 in accordance with these statements in the RFC:
1141 The auto-replied keyword:
1143 - SHOULD be used on messages sent in direct response to another
1144 message by an automatic process,
1146 - MUST NOT be used on manually-generated messages,
1148 - MAY be used on Delivery Status Notifications (DSNs) and Message
1149 Disposition Notifications (MDNs),
1151 - MUST NOT be used on messages generated by automatic or periodic
1152 processes, except for messages which are automatic responses to
1155 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1156 to the default Received: header definition.
1158 PH/12 Added log selector acl_warn_skipped (default on).
1160 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1161 variables because (a) they are in the wrong storage pool and (b) even if
1162 they were copied, it wouldn't work properly because of the caching.
1164 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1165 checking when decoding. Apparently there are clients that generate
1166 overlong encoded strings. Why am I not surprised?
1168 PH/15 If the first argument of "${if match_address" was not empty, but did not
1169 contain an "@" character, Exim crashed. Now it writes a panic log message
1170 and treats the condition as false.
1172 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1174 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1175 extension "envelope-auth". The code is finished and in agreement with
1176 other implementations, but there is no documentation so far and in fact,
1177 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1178 not changing the active code.
1180 Print executed "if" and "elsif" statements when debugging is used. This
1181 helps a great deal to understand what a filter does.
1183 Document more things not specified clearly in RFC3028. I had all this
1184 sorted out, when out of a sudden new issues came to my mind. Oops."
1186 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1189 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1190 canonical form (as documented). However, after a host name lookup from
1191 the IP address, check_host() was doing a simple string comparison with
1192 addresses acquired from the DNS when checking that the found name did
1193 have the original IP as one of its addresses. Since any found IPv6
1194 addresses are likely to be in abbreviated form, the comparison could
1195 fail. Luckily, there already exists a function for doing the comparison
1196 by converting both addresses to binary, so now that is used instead of
1197 the text comparison.
1199 PH/20 There was another similar case to PH/19, when a complete host name was
1200 given in a host list; looking up its IP address could give an abbreviated
1201 form, whereas the current host's name might or might not be abbreviated.
1202 The same fix has been applied.
1208 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1209 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1212 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1213 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1215 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1217 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1218 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1219 Also, for "Auto-Submitted", treat anything other than "no" as
1220 non-personal, in accordance with RFC 3834. (Previously it treated
1221 anything starting "auto-" as non-personal.)
1223 TF/01 The control=submission/name=... option had a problem with syntax
1224 errors if the name included a slash character. The /name= option
1225 now slurps the rest of the string, so it can include any characters
1226 but it must come last in the list of options (after /sender_retain
1229 PH/05 Some modifications to the interface to the fake nameserver for the new
1237 TK/01 Added the "success_on_redirect" address verification option. See
1238 NewStuff for rationale and an example.
1240 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1242 PH/02 Patch to exigrep to allow it to work on syslog lines.
1244 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1245 fread() to skip over the body file's header line, because in Cygwin the
1246 header line is locked and is inaccessible.
1248 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1249 co-exist for some time) to make it clear that it is the Exim ID that is
1250 referenced, not the Message-ID: header line.
1252 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1253 string_format() function, because snprintf() does not exist on all
1256 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1257 ${stat: expansion item.
1259 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1260 protocol synchronization error", to keep the pedants happy.
1262 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1263 well as for IRIX systems, when gcc is being used. See the host.c source
1266 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1268 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1271 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1272 transport and to the smtp transport in LMTP mode.
1274 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1276 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1277 run a filter in a subprocess. This could lead to confusion in subsequent
1278 lookups in the parent process. There should also be a search_tidyup() at
1279 the end of the subprocess.
1281 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1282 only if the host matched helo_try_verify_hosts, which caused the
1283 verification to occur when the EHLO/HELO command was issued. The ACL just
1284 tested the remembered result. Now, if a previous verification attempt has
1285 not happened, "verify = helo" does it there and then.
1287 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1289 TK/03 Fix log output including CR from clamd.
1291 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1292 exist provoked a memory error which could cause a segfault.
1294 PH/15 Installed PCRE 6.2
1296 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1298 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1299 of the problem. Specifically, suggested +O2 rather than +O1 for the
1302 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1304 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1305 record, the port was not correctly logged when the outgoing_port log
1306 selector was set (it logged the transort's default port).
1308 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1309 fallback_hosts, and "hosts" in the smtp transport.
1311 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1312 host errors such as "Connection refused".
1314 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1315 authentication with radiusclient 0.4.9:
1317 - Error returned from rc_read_config was caught wrongly
1318 - Username/password not passed on to radius server due to wrong length.
1320 The presumption is that some radiusclient API changes for 4.51/PH/17
1321 were not taken care of correctly. The code is still untested by me (my
1322 Linux distribution still has 0.3.2 of radiusclient), but it was
1323 contributed by a Radius user.
1325 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1326 expanding the "port" option of the smtp transport.
1328 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1329 while reading a MIME header. Thanks to Tom Hughes for a patch.
1331 PH/24 Include config.h inside local_scan.h so that configuration settings are
1334 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1335 that in an ACL for RCPT (for example), you can examine exactly what was
1338 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1339 commands, but it was not correctly comparing the address with the actual
1340 client host address. Thus, it would show the EHLO address in Received:
1341 header lines when this was not necessary.
1343 PH/27 Added the % operator to ${eval:}.
1345 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1346 it should be ignoring failures (because with -C, for example, it has lost
1347 privilege). It wasn't ignoring creation failures other than "already
1350 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1353 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1354 because an input line was too long, either on its own, or by virtue of
1355 too many continuations, the temporary file was not being removed, and the
1356 return code was incorrect.
1358 PH/31 Missing "BOOL" in function definition in filtertest.c.
1360 PH/32 Applied Sieve patches from the maintainer.
1362 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1364 PH/33 Added "verify = not_blind".
1366 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1367 Local/Makefile (with some defaults set). These are used in built scripts
1368 such as exicyclog, but they have never been used in the exim_install
1369 script (though there are many overriding facilities there). I have
1370 arranged that the exim_install script now takes note of these two
1373 PH/35 Installed configuration files for Dragonfly.
1375 PH/36 When a locally submitted message by a trusted user did not contain a
1376 From: header, and the sender address was obtained from -f or from an SMTP
1377 MAIL command, and the trusted user did not use -F to supply a sender
1378 name, $originator_name was incorrectly used when constructing a From:
1379 header. Furthermore, $originator_name was used for submission mode
1380 messages from external hosts without From: headers in a similar way,
1381 which is clearly wrong.
1383 PH/37 Added control=suppress_local_fixups.
1385 PH/38 When log_selector = +received_sender was set, and the addition of the
1386 sender made the log line's construction buffer exactly full, or one byte
1387 less than full, an overflow happened when the terminating "\n" was
1390 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1391 when the result of a list match is failure because a DNS lookup failed.
1393 PH/40 RM_COMMAND is now used in the building process.
1395 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1396 the "build-* directories that it finds.
1398 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1399 address in a domain literal was a prefix of an interface address.
1401 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1402 when verifying a sender address, unless rewrite_headers is false.
1404 PH/44 Wrote a long comment about why errors_to addresses are verified as
1405 recipients, not senders.
1407 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1408 the ratelimit ACL was added.
1410 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1412 PH/46 Added extra information about PostgreSQL errors to the error string.
1414 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1415 avoiding the need to install special zones in a real server. This is
1416 backwards compatible; if it can't find the fake resolver, it drops back.
1417 Thus, both old and new test suites can be run.
1419 TF/02 Added util/ratelimit.pl
1421 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1422 clock is set back in time.
1424 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1425 Candler <B.Candler@pobox.com>.
1427 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1428 OK for addresses that are the result of redirections.
1430 PH/48 A number of further additions for the benefit of the new test suite,
1431 including a fake gethostbyname() that interfaces to the fake DNS resolver
1434 TF/06 The fix for widen_domains has also been applied to qualify_single and
1435 search_parents which are the other dnslookup options that can cause
1438 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1441 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1443 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1444 service" if it encountered an MX record with an empty target hostname.
1445 The message is now "an MX or SRV record indicated no SMTP service".
1447 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1448 if the DNS of the sending site is misconfigured. This is quite a
1449 common situation. This change restores the behaviour of treating a
1450 helo verification defer as a failure.
1452 PH/51 If self=fail was set on a router, the bounce message did not include the
1453 actual error message.
1459 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1461 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1462 command itself ended in error, the underlying message about the transport
1463 filter timeout was being overwritten with the pipe command error. Now the
1464 underlying error message should be appended to the second error message.
1466 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1468 PH/02 Reduce the amount of output that "make" produces by default. Full output
1469 can still be requested.
1471 PH/03 The warning log line about a condition test deferring for a "warn" verb
1472 was being output only once per connection, rather than after each
1473 occurrence (because it was using the same function as for successful
1474 "warn" verbs). This seems wrong, so I have changed it.
1476 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1477 it should not have, which might have caused a crash in the right
1478 circumstances, but probably never did.
1480 PH/04 Installed a modified version of Tony Finch's patch to make submission
1481 mode fix the return path as well as the Sender: header line, and to
1482 add a /name= option so that you can make the user's friendly name appear
1485 TF/03 Added the control = fakedefer ACL modifier.
1487 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1488 Mark Lowes for thorough testing.
1490 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1492 TK/03 Merged latest SRS patch from Miles Wilton.
1494 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1495 with the definition in sysexits.h (which is #included earlier).
1496 Fortunately, Exim does not actually use EX_OK. The code used to try to
1497 preserve the sysexits.h value, by assumimg that macro definitions were
1498 scanned for macro replacements. I have been disabused of this notion,
1499 so now the code just undefines EX_OK before #including unistd.h.
1501 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1502 in the smtp transport. When a block could not be written in a single
1503 write() function, the timeout was being re-applied to each part-write.
1504 This seems wrong - if the receiver was accepting one byte at a time it
1505 would take for ever. The timeout is now adjusted when this happens. It
1506 doesn't have to be particularly precise.
1508 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1509 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1511 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1512 without a domain if the check to <postmaster@domain> fails.
1514 SC/01 Eximstats: added -xls and the ability to specify output files
1515 (patch written by Frank Heydlauf).
1517 SC/02 Eximstats: use FileHandles for outputing results.
1519 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1521 SC/04 Eximstats: fixed display of large numbers with -nvr option
1523 SC/05 Eximstats: fixed merging of reports with empty tables.
1525 SC/06 Eximstats: added the -include_original_destination flag
1527 SC/07 Eximstats: removed tabs and trailing whitespace.
1529 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1531 TK/06 MBOX spool code: Add real "From " MBOX separator line
1532 so the .eml file is really in mbox format (even though
1533 most programs do not really care). Patch from Alex Miller.
1535 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1536 The latter is generated from $received_to and is only set if the
1537 message has one envelope recipient. SA can use these headers,
1538 obviously out-of-the-box. Patch from Alex Miller.
1540 PH/08 The ${def test on a variable was returning false if the variable's
1541 value was "0", contrary to what the specification has always said!
1542 The result should be true unless the variable is empty.
1544 PH/09 The syntax error of a character other than { following "${if
1545 def:variable_name" (after optional whitespace) was not being diagnosed.
1546 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1547 accidental colon was present, for example, could give incorrect results.
1549 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1550 result is used (not including appendfile, where other changes are about
1553 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1554 This involved changing a lot of size variables from int to off_t. It
1555 should work with maildirs and everything.
1557 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1558 spamd dying while we are connected to it.
1560 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1563 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1564 function for BDB 4.3.
1566 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1568 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1569 the sources for PCRE have been re-arranged and more files are now
1572 PH/15 The code I had for printing potentially long long variables in PH/11
1573 above was not the best (it lost precision). The length of off_t variables
1574 is now inspected at build time, and an appropriate printing format (%ld
1575 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1576 to be "long long int" or "long int". This is needed for the internal
1577 formatting function string_vformat().
1579 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1580 the configuration file to be ":syslog", then the script "guesses" where
1581 the logs files are, rather than using the compiled in default. In our
1582 case the guess is not the same as the compiled default, so the script
1583 suddenly stopped working when I started to use syslog. The patch checks
1584 to see if log_file_path is "". If so, it attempts to read it from exim
1585 with no configuration file to get the compiled in version, before it
1586 falls back to the previous guessing code."
1588 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1589 implementing BATV in an Exim configuration. See NewStuff for the gory
1592 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1593 Makefile that are specific to HP-UX.
1595 PH/18 If the "use_postmaster" option was set for a recipient callout together
1596 with the "random" option, the postmaster address was used as the MAIL
1597 FROM address for the random test, but not for the subsequent recipient
1598 test. It is now used for both.
1600 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1601 patch removes a few documentation additions to RFC 3028, because the
1602 latest draft now contains them. It adds the new en;ascii-case comparator
1603 and a new error check for 8bit text in MIME parts. Comparator and
1604 require names are now matched exactly. I enabled the subaddress
1605 extension, but it is not well tested yet (read: it works for me)."
1607 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1608 rework some of the code of TK/09 above to avoid the hardwired use of
1609 "%lld" and "long long". Replaced the call to snprintf() with a call to
1612 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1613 records point to non-existent hosts", "retry timeout exceeded", and
1614 "retry time not reached for any host after a long failure period".
1616 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1617 experimental DomainKeys support:
1619 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1620 (2) On an error such as an illegally used "control", the wrong name for
1621 the control was given.
1623 These problems did NOT occur unless DomainKeys support was compiled.
1625 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1627 PH/24 Added ${if match_ip condition.
1629 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1630 databases so that it will be absolutely obvious if a crash occurs in the
1631 DB library. This is a regular occurrence (often caused by mis-matched
1634 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1635 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1636 picked up on a user's system that detects such things. There doesn't seem
1637 to be a gcc warning option for this - only an attribute that has to be
1638 put on the function's prototype. It seems that in Fedora Core 4 they have
1639 set this on a number of new functions. No doubt there will be more in due
1642 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1643 specify a transport. However, if an address that was verified by such a
1644 router was the subject of a callout, Exim crashed because it tried to
1645 read the rcpt_include_affixes from the non-existent transport. Now it
1646 just assumes that the setting of that option is false. This bug was
1647 introduced by 4.51/PH/31.
1649 PH/28 Changed -d+all to exclude +memory, because that information is very
1650 rarely of interest, but it makes the output a lot bigger. People tend to
1651 do -d+all out of habit.
1653 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1654 code in os-type was giving problems when libc.so lives in lib64, like on
1657 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1658 aren't the modern standard, and it seems that some systems' include files
1659 don't always have them. Exim was already checking for some of the newer
1660 ones like T_AAAA, and defining it itself. I've added checks for all the
1661 record types that Exim uses.
1663 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1664 not automatically generating a new one, as it is supposed to. This
1665 prevented TLS from working. If the file did exist, but contained invalid
1666 data, a new version was generated, as expected. It was only the case of a
1667 non-existent file that was broken.
1669 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1670 with a change in libdomainkeys > 0.64.
1672 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1673 from DNS. If the selector record carries the flag, it now has
1674 precedence over the domain-wide flag.
1676 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1678 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1679 the use of a port name that isn't defined in /etc/services) occurred, the
1680 message was deferred as in a normal delivery, and thus remained on the
1681 spool, instead of being failed because of the mua_wrapper setting. This
1682 is now fixed, and I tidied up some of the mua_wrapper messages at the
1685 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1686 the messages in a hash of arrays rather than using individual hashes.
1687 This is a bit cleaner and results in dramatic memory savings, albeit
1688 at a slight CPU cost.
1690 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1691 as requested by Marc Sherman.
1693 SC/10 Eximstats: added histograms for user specified patterns as requested
1696 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1698 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1699 fopen() in the content-scanning modules that did not already have it.
1705 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1706 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1708 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1710 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1712 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1713 to test Sieve filters that use "vacation".
1715 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1716 that changes the way the GnuTLS parameters are stored in the cache file.
1717 The new format can be generated externally. For backward compatibility,
1718 if the data in the cache doesn't make sense, Exim assumes it has read an
1719 old-format file, and it generates new data and writes a new file. This
1720 means that you can't go back to an older release without removing the
1723 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1724 work if there are any delivery delays because "one_time" forces the
1725 parent to be marked "delivered", so its unseen clone is never tried
1726 again. For this reason, Exim now forbids the simultaneous setting of
1729 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1730 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1731 entry if you want to know the details. The fix, however, neglected to
1732 consider the case when local delivery batching is involved. The test for
1733 "previously delivered" was not happening when checking to see if an
1734 address could be batched with a previous (undelivered) one; under
1735 certain circumstances this could lead to multiple deliveries to the same
1738 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1739 in its include files, and this causes problems building Exim.
1741 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1742 header_syntax) but Exim was just ignoring anything given after a slash.
1743 In particular, this caused confusion with an attempt to use "verify =
1744 reverse_host_lookup/defer_ok". An error is now given when options are
1745 supplied for verify items that do not have them. (Maybe reverse_host_
1746 lookup should have a defer_ok option, but that's a different point.)
1748 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1749 defined by RFC 821) to 2048, because there were problems with some AUTH
1750 commands, and RFC 1869 says the size should be increased for extended
1751 SMTP commands that take arguments.
1753 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1756 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1757 "unknown" error; now it says that the functionality isn't in the binary.
1759 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1760 an address' error message when a string expansion fails (syntax or
1761 whatever). Otherwise the password may appear in the log. Following change
1762 PH/42 below, there is no longer a chance of it appearing in a bounce
1765 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1767 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1768 of its addresses was ever tried. (Bugzilla bug #2.)
1770 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1771 the result incorrectly in the debug output. (It correctly added a newline
1772 to what was transported.)
1774 TF/01 Added $received_time.
1776 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1777 commented out examples of how to interface to a virus scanner and to
1778 SpamAssassin. Also added commented examples of av_scanner and
1779 spamd_address settings.
1781 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1782 and controls are allowed in which ACLs. There were a couple of minor
1783 errors. Some of the entries in the conditions table (which is a table of
1784 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1785 negation of where the condition IS allowed.
1787 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1789 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1790 header file does not have a version number, so I've had to invent a new
1791 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1792 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1793 radiusclient), but it was contributed by a Radius user.
1795 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1796 files or directories, for OpenSSL.
1798 PH/19 When an Exim process that is running as root has to create an Exim log
1799 file, it does so in a subprocess that runs as exim:exim so as to get the
1800 ownership right at creation (otherwise, other Exim processes might see
1801 the file with the wrong ownership). There was no test for failure of this
1802 fork() call, which would lead to the process getting stuck as it waited
1803 for a non-existent subprocess. Forks do occasionally fail when resources
1804 run out. I reviewed all the other calls to fork(); they all seem to check
1807 PH/20 When checking for unexpected SMTP input at connect time (before writing
1808 the banner), Exim was not dealing correctly with a non-positive return
1809 from the read() function. If the client had disconnected by this time,
1810 the result was a log entry for a synchronization error with an empty
1811 string after "input=" when read() returned zero. If read() returned -1
1812 (an event I could not check), uninitialized data bytes were printed.
1813 There were reports of junk text (parts of files, etc) appearing after
1816 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1818 PH/22 Added support for macro redefinition, and (re)definition in between
1819 driver and ACL definitions.
1821 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1822 forgetting to use the resulting value; it was using the unexpanded value.
1824 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1825 hadn't been configured. The fix is from Juergen Kreileder, who
1826 understands it better than I do:
1828 "Here's what I see happening with three configured cyrus_sasl
1829 authenticators configured (plain, login, cram-md5):
1831 On startup auth_cyrus_sasl_init() gets called for each of these.
1832 This means three calls to sasl_listmech() without a specified mech_list.
1833 => SASL tests which mechs of all available mechs actually work
1834 => three warnings about OTP not working
1835 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1837 With the patch, sasl_listmech() also gets called three times. But now
1838 SASL's mech_list option is set to the server_mech specified in the the
1839 authenticator. Or in other words, the answer from sasl_listmech()
1840 gets limited to just the mech you're testing for (which is different
1842 => the return list contains just 'plain' or 'login', 'cram-md5' or
1843 nothing depending on the value of ob->server_mech.
1845 I've just tested the patch: Authentication still works fine,
1846 unavailable mechs specified in the exim configuration are still
1847 caught, and the auth.log warnings about OTP are gone."
1849 PH/25 When debugging is enabled, the contents of the command line are added
1850 to the debugging output, even when log_selector=+arguments is not
1853 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1854 answer is "GNU", and only if the return is "GNU/something" is the answer
1857 PH/27 $acl_verify_message is now set immediately after the failure of a
1858 verification in an ACL, and so is available in subsequent modifiers. In
1859 particular, the message can be preserved by coding like this:
1861 warn !verify = sender
1862 set acl_m0 = $acl_verify_message
1864 Previously, $acl_verify_message was set only while expanding "message"
1865 and "log_message" when a very denied access.
1867 PH/28 Modified OS/os.c-Linux with
1869 -#ifndef OS_LOAD_AVERAGE
1870 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1872 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1873 nomenclature these days.)
1875 PH/29 Installed patch from the Sieve maintainer that adds the options
1876 sieve_useraddress and sieve_subaddress to the redirect router.
1878 PH/30 In these circumstances:
1879 . Two addresses routed to the same list of hosts;
1880 . First host does not offer TLS;
1881 . First host accepts first address;
1882 . First host gives temporary error to second address;
1883 . Second host offers TLS and a TLS session is established;
1884 . Second host accepts second address.
1885 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1886 and peerdn, if requested) that were in fact used only for the second
1889 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1890 attention to any local part prefix or suffix that was matched by the
1891 router that accepted the address. It now behaves in the same way as it
1892 does for delivery: the affixes are removed from the local part unless
1893 rcpt_include_affixes is set on the transport.
1895 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1896 timeout during the DATA phase of an incoming message.
1898 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1899 applied a patch sanctioned by the Sieve maintainer.
1901 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1902 the uid or gid is negative. A case of a negative gid caused this to be
1903 noticed. The fix allows for either to be negative.
1905 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1906 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1909 PH/36 The change PH/12 above was broken. Fixed it.
1911 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1912 the grounds that routing the same address twice would always produce the
1913 same answer. This might have been true once, but it is certainly no
1914 longer true now. Routing a child address may depend on the previous
1915 routing that produced that child. Some complicated redirection strategies
1916 went wrong when messages had multiple recipients, and made Exim's
1917 behaviour dependent on the order in which the addresses were given.
1919 I have moved the duplicate checking until after the routing is complete.
1920 Exim scans the addresses that are assigned to local and remote
1921 transports, and removes any duplicates. This means that more work will be
1922 done, as duplicates will always all be routed, but duplicates are
1923 presumably rare, so I don't expect this is of any significance.
1925 For deliveries to pipes, files, and autoreplies, the duplicate checking
1926 still happens during the routing process, since they are not going to be
1929 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1930 It corrects a timeout issue with spamd. This is Ian's comment: "The
1931 background is that sometimes spamd either never reads data from a
1932 connection it has accepted, or it never writes response data. The exiscan
1933 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1934 blindly assumes that writes won't block so it may never time out."
1936 PH/39 Allow G after quota size as well as K and M.
1938 PH/40 The value set for $authenticated_id in an authenticator may not contain
1939 binary zeroes or newlines because the value is written to log lines and
1940 to spool files. There was no check on this. Now the value is run through
1941 the string_printing() function so that such characters are converted to
1942 printable escape sequences.
1944 PH/41 $message_linecount is a new variable that contains the total number of
1945 lines in the message. Compare $body_linecount, which is the count for the
1948 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1949 bounce and delay warning messages, except in certain special cases, which
1952 (a) An SMTP error message from a remote host;
1953 (b) A message specified in a :fail: redirection;
1954 (c) A message specified in a "fail" command in a system filter;
1955 (d) A message specified in a FAIL return from the queryprogram router;
1956 (e) A message specified by the cannot_route_message router option.
1958 In these cases only, Exim does include the error details in bounce and
1959 warning messages. There are also a few cases where bland messages such
1960 as "unrouteable address" or "local delivery error" are given.
1962 PH/43 $value is now also set for the "else" part of a ${run expansion.
1964 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1965 being worked on, but at least Exim now implements the latest version to
1968 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1969 process to complete was treated as a delivery failure, a timeout while
1970 writing the message to the pipe was logged, but erroneously treated as a
1971 successful delivery. Such timeouts include transport filter timeouts. For
1972 consistency with the overall process timeout, these timeouts are now
1973 treated as errors, giving rise to delivery failures by default. However,
1974 there is now a new Boolean option for the pipe transport called
1975 timeout_defer, which, if set TRUE, converts the failures into defers for
1976 both kinds of timeout. A transport filter timeout is now identified in
1979 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1980 systems where "make" and "gmake" are different, calling "gmake" at top
1981 level broke things. I've arranged for the value of $(MAKE) to be passed
1982 from the Makefile to this script so that it can call the same version of
1986 A note about Exim versions 4.44 and 4.50
1987 ----------------------------------------
1989 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1990 changes of various kinds. As a consequence, a big documentation update was
1991 needed. This delayed the release for rather longer than seemed good, especially
1992 in the light of a couple of (minor) security issues. Therefore, the changes
1993 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1994 release. So 4.44 and 4.50 are in effect two different branches that both start
1997 I have left the 4.50 change log unchanged; it contains all the changes since
1998 4.43. The change log for 4.44 is below; many of its items are identical to
1999 those for 4.50. This seems to be the most sensible way to preserve the
2000 historical information.
2006 1. Minor wording change to the doc/README.SIEVE file.
2008 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2009 computation of the current number of files was incorrect.
2011 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2012 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2015 4. Give more explanation in the error message when the command for a transport
2016 filter fails to execute.
2018 5. There are several places where Exim runs a non-Exim command in a
2019 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2020 was being done only for the command run by the queryprogram router. It is
2021 now done for all such subprocesses. The other cases are: ${run, transport
2022 filters, and the commands run by the lmtp and pipe transports.
2024 6. Added CONFIGURE_GROUP build-time option.
2026 7. Some older OS have a limit of 256 on the maximum number of file
2027 descriptors. Exim was using setrlimit() to set 1000 as a large value
2028 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2029 systems. I've change it so that if it can't get 1000, it tries for 256.
2031 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2032 was an oversight, and furthermore, ever since the addition of extra
2033 controls (e.g. 4.43/32), the checks on when to allow different forms of
2034 "control" were broken. There should now be diagnostics for all cases when a
2035 control that does not make sense is encountered.
2037 9. Added the /retain_sender option to "control=submission".
2039 10. $recipients is now available in the predata ACL (oversight).
2041 11. Tidy the search cache before the fork to do a delivery from a message
2042 received from the command line. Otherwise the child will trigger a lookup
2043 failure and thereby defer the delivery if it tries to use (for example) a
2044 cached ldap connection that the parent has called unbind on.
2046 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2047 of $address_data from the recipient verification was clobbered by the
2048 sender verification.
2050 13. The value of address_data from a sender verification is now available in
2051 $sender_address_data in subsequent conditions in the ACL statement.
2053 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2055 15. Added a new option "connect=<time>" to callout options, to set a different
2058 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2059 was its contents. (It was OK if the option was not defined at all.)
2061 17. A "Completed" log line is now written for messages that are removed from
2062 the spool by the -Mrm option.
2064 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2065 information about exactly what failed.
2067 19. Added -dd to debug only the daemon process.
2069 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2070 handles timeouts, both on the server side and network timeouts. Renamed the
2071 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2073 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2074 It is now set to "smtps".
2076 22. $host_address is now set to the target address during the checking of
2077 ignore_target_hosts.
2079 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2080 being passed; this would have caused $sender_host_name to have been used if
2081 matching the list had actually called for a host name (not very likely,
2082 since this list is usually IP addresses). A host name is now passed as
2085 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2086 code that specifies a non-restarting handler (typically sigaction() in
2087 modern systems) in an attempt to fix a rare and obscure crash bug.
2089 25. Narrowed the window for a race in the daemon that could cause it to ignore
2090 SIGCHLD signals. This is not a major problem, because they are used only to
2091 wake it up if nothing else does.
2093 26. A malformed maildirsize file could cause Exim to calculate negative values
2094 for the mailbox size or file count. Odd effects could occur as a result.
2095 The maildirsize information is now recalculated if the size or filecount
2098 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2099 support for a long time. Removed HAVE_SYS_VFS_H.
2101 28. Installed the latest version of exipick from John Jetmore.
2103 29. In an address list, if the pattern was not a regular expression, an empty
2104 subject address (from a bounce message) matched only if the pattern was an
2105 empty string. Non-empty patterns were not even tested. This was the wrong
2106 because it is perfectly reasonable to use an empty address as part of a
2107 database query. An empty address is now tested by patterns that are
2108 lookups. However, all the other forms of pattern expect the subject to
2109 contain a local part and a domain, and therefore, for them, an empty
2110 address still always fails if the pattern is not itself empty.
2112 30. Exim went into a mad DNS loop when attempting to do a callout where the
2113 host was specified on an smtp transport, and looking it up yielded more
2114 than one IP address.
2116 31. Re-factored the code for checking spool and log partition space into a
2117 function that finds that data and another that does the check. The former
2118 is then used to implement four new variables: $spool_space, $log_space,
2119 $spool_inodes, and $log_inodes.
2121 32. The RFC2047 encoding function was originally intended for short strings
2122 such as real names; it was not keeping to the 75-character limit for
2123 encoded words that the RFC imposes. It now respects the limit, and
2124 generates multiple encoded words if necessary. To be on the safe side, I
2125 have increased the buffer size for the ${rfc2047: expansion operator from
2128 33. It is now permitted to omit both strings after an "if" condition; if the
2129 condition is true, the result is "true". As before, when the second string
2130 is omitted, a false condition yields an empty string. This makes it less
2131 cumbersome to write custom ACL and router conditions.
2133 34. Failure to deliver a bounce message always caused it to be frozen, even if
2134 there was an errors_to setting on the router. The errors_to setting is now
2137 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2138 canonical form (fully expanded) before being placed in
2139 $sender_host_address.
2141 36. The table in the code that translates DNS record types into text (T_A to
2142 "A" for instance) was missing entries for NS and CNAME. It is just possible
2143 that this could have caused confusion if both these types were looked up
2144 for the same domain, because the text type is used as part of Exim's
2145 per-process caching. But the chance of anyone hitting this buglet seems
2148 37. The dnsdb lookup has been extended in a number of ways.
2150 (1) There is a new type, "zns", which walks up the domain tree until it
2151 finds some nameserver records. It should be used with care.
2153 (2) There is a new type, "mxh", which is like "mx" except that it returns
2154 just the host names, not the priorities.
2156 (3) It is now possible to give a list of domains (or IP addresses) to be
2157 looked up. The behaviour when one of the lookups defers can be
2158 controlled by a keyword.
2160 (4) It is now possible to specify the separator character for use when
2161 multiple records are returned.
2163 38. The dnslists ACL condition has been extended: it is now possible to supply
2164 a list of IP addresses and/or domains to be looked up in a particular DNS
2167 39. Added log_selector=+queue_time_overall.
2169 40. When running the queue in the test harness, wait just a tad after forking a
2170 delivery process, to get repeatability of debugging output.
2172 41. Include certificate and key file names in error message when GnuTLS fails
2173 to set them up, because the GnuTLS error message doesn't include the name
2174 of the failing file when there is a problem reading it.
2176 42. Allow both -bf and -bF in the same test run.
2178 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2180 44. The "Exiscan patch" is now merged into the mainline Exim source.
2182 45. Sometimes the final signoff response after QUIT could fail to get
2183 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2184 < 0 before doing a fflush(). This bug looks as though it goes back to the
2185 introduction of TLS in release 3.20, but "sometimes" must have been rare
2186 because the tests only now provoked it.
2188 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2189 (this can affect the format of dates).
2191 47. exim_tidydb, when checking for the continued existence of a message for
2192 which it has found a message-specific retry record, was not finding
2193 messages that were in split spool directories. Consequently, it was
2194 deleting retry records that should have stayed in existence.
2196 48. Steve fixed some bugs in eximstats.
2198 49. The SPA authentication driver was not abandoning authentication and moving
2199 on to the next authenticator when an expansion was forced to fail,
2200 contradicting the general specification for all authenticators. Instead it
2201 was generating a temporary error. It now behaves as specified.
2203 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2204 (the order specifies the preference for clients). The order is now AES256,
2205 AES128, 3DES, ARCFOUR128.
2207 51. Small patch to Sieve code - explicitly set From: when generating an
2210 52. Exim crashed if a remote delivery caused a very long error message to be
2211 recorded - for instance if somebody sent an entire SpamAssassin report back
2212 as a large number of 550 error lines. This bug was coincidentally fixed by
2213 increasing the size of one of Exim's internal buffers (big_buffer) that
2214 happened as part of the Exiscan merge. However, to be on the safe side, I
2215 have made the code more robust (and fixed the comments that describe what
2218 53. Now that there can be additional text after "Completed" in log lines (if
2219 the queue_time_overall log selector is set), a one-byte patch to exigrep
2220 was needed to allow it to recognize "Completed" as not the last thing in
2223 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2224 patch that reportedly fixes this has been added. I am not expert enough to
2225 create a test for it. This is what the patch creator wrote:
2227 "I found a little strange behaviour of ldap code when working with
2228 Windows 2003 AD Domain, where users was placed in more than one
2229 Organization Units. When I tried to give exim partial DN, the exit code
2230 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2231 But simultaneously result of request was absolutely normal ldap result,
2232 so I produce this patch..."
2234 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2235 REFERENCE, so I have modified the code to exclude the patch when that macro
2238 55. Some experimental protocols are using DNS PTR records for new purposes. The
2239 keys for these records are domain names, not reversed IP addresses. The
2240 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2241 leaves it alone. Component reversal etc. now happens only for IP addresses.
2244 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2246 57. Double the size of the debug message buffer (to 2048) so that more of very
2247 long debug lines gets shown.
2249 58. The exicyclog utility now does better if the number of log files to keep
2250 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2252 59. Two changes related to the smtp_active_hostname option:
2254 (1) $smtp_active_hostname is now available as a variable.
2255 (2) The default for smtp_banner uses $smtp_active_hostname instead
2256 of $primary_hostname.
2258 60. The host_aton() function is supposed to be passed a string that is known
2259 to be a valid IP address. However, in the case of IPv6 addresses, it was
2260 not checking this. This is a hostage to fortune. Exim now panics and dies
2261 if the condition is not met. A case was found where this could be provoked
2262 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2263 components; fortuitously, this particular loophole had already been fixed
2264 by change 4.50/55 above.
2266 If there are any other similar loopholes, the new check in host_aton()
2267 itself should stop them being exploited. The report I received stated that
2268 data on the command line could provoke the exploit when Exim was running as
2269 exim, but did not say which command line option was involved. All I could
2270 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2271 running as the user.
2274 61. There was a buffer overflow vulnerability in the SPA authentication code
2275 (which came originally from the Samba project). I have added a test to the
2276 spa_base64_to_bits() function which I hope fixes it.
2279 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2280 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2282 63. The daemon start-up calls getloadavg() while still root for those OS that
2283 need the first call to be done as root, but it missed one case: when
2284 deliver_queue_load_max is set with deliver_drop_privilege. This is
2285 necessary for the benefit of the queue runner, because there is no re-exec
2286 when deliver_drop_privilege is set.
2288 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2289 This has been fixed.
2291 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2292 was in use, was not putting the data itself into the right store pool;
2293 consequently, it could be overwritten for a subsequent message in the same
2294 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2297 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2299 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2300 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2301 the function were treating the return as a boolean value, which happened to
2302 work because 0=false and not-0=true, but is not correct code.
2304 68. The host_aton() function was not handling scoped IPv6 addresses (those
2305 with, for example, "%eth0" on the end) correctly.
2307 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2308 negated items (that is, ~something) in unsigned ints. Some compilers
2309 apparently mutter when there is no cast.
2311 70. If an address verification called from an ACL failed, and did not produce a
2312 user-specific message (i.e. there was only a "system" message), nothing was
2313 put in $acl_verify_message. In this situation, it now puts the system
2316 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2317 session; change 4.31/43 added the unwanted input to the log line - except
2318 that it did not do this in the start of session case. It now does.
2320 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2321 This is wrong and can cause the other end to generate a synchronization
2322 error if it is another Exim or anything else that does the synchronization
2323 check. A QUIT command is no longer sent after a timeout.
2325 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2326 during host lookups.
2328 74. The defer_ok option of callout verification was not working if it was used
2329 when verifying addresses in header lines, that is, for this case:
2331 verify = header_sender/callout=defer_ok
2333 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2334 those file descriptors could be used for SMTP connections. If anything
2335 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2336 could be sent to the SMTP client, causing chaos. The daemon now opens
2337 stdin, stdout, and stderr to /dev/null when it puts itself into the
2340 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2341 log by default. The user can override this with suitable Perl magic.
2343 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2344 the log message when discard triggers, was not working for the DATA ACL or
2345 for the non-SMTP ACL.
2347 78. Error message wording change in sieve.c.
2349 79. If smtp_accept_max_per_host was set, the number of connections could be
2350 restricted to fewer than expected, because the daemon was trying to set up
2351 a new connection before checking whether the processes handling previous
2352 connections had finished. The check for completed processes is now done
2353 earlier. On busy systems, this bug wouldn't be noticed because something
2354 else would have woken the daemon, and it would have reaped the completed
2357 80. If a message was submitted locally by a user whose login name contained one
2358 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2359 It caused a spool format error. I have fixed the spool reading code. A
2360 related problem was that the "from" clause in the Received: line became
2361 illegal because of the space(s). It is now covered by ${quote_local_part.
2363 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2366 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2368 83. Patch from Sieve maintainer to fix unterminated string problem in
2369 "vacation" handling.
2371 84. Some minor changes to the Linux configuration files to help with other
2372 OS variants using glibc.
2374 85. One more patch for Sieve to update vacation handling to latest spec.
2377 ----------------------------------------------------
2378 See the note above about the 4.44 and 4.50 releases.
2379 ----------------------------------------------------
2385 1. Change 4.43/35 introduced a bug that caused file counts to be
2386 incorrectly computed when quota_filecount was set in an appendfile
2389 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2390 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2393 3. Give more explanation in the error message when the command for a transport
2394 filter fails to execute.
2396 4. There are several places where Exim runs a non-Exim command in a
2397 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2398 was being done only for the command run by the queryprogram router. It is
2399 now done for all such subprocesses. The other cases are: ${run, transport
2400 filters, and the commands run by the lmtp and pipe transports.
2402 5. Some older OS have a limit of 256 on the maximum number of file
2403 descriptors. Exim was using setrlimit() to set 1000 as a large value
2404 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2405 systems. I've change it so that if it can't get 1000, it tries for 256.
2407 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2408 was an oversight, and furthermore, ever since the addition of extra
2409 controls (e.g. 4.43/32), the checks on when to allow different forms of
2410 "control" were broken. There should now be diagnostics for all cases when a
2411 control that does not make sense is encountered.
2413 7. $recipients is now available in the predata ACL (oversight).
2415 8. Tidy the search cache before the fork to do a delivery from a message
2416 received from the command line. Otherwise the child will trigger a lookup
2417 failure and thereby defer the delivery if it tries to use (for example) a
2418 cached ldap connection that the parent has called unbind on.
2420 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2421 of $address_data from the recipient verification was clobbered by the
2422 sender verification.
2424 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2425 was its contents. (It was OK if the option was not defined at all.)
2427 11. A "Completed" log line is now written for messages that are removed from
2428 the spool by the -Mrm option.
2430 12. $host_address is now set to the target address during the checking of
2431 ignore_target_hosts.
2433 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2434 being passed; this would have caused $sender_host_name to have been used if
2435 matching the list had actually called for a host name (not very likely,
2436 since this list is usually IP addresses). A host name is now passed as
2439 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2440 code that specifies a non-restarting handler (typically sigaction() in
2441 modern systems) in an attempt to fix a rare and obscure crash bug.
2443 15. Narrowed the window for a race in the daemon that could cause it to ignore
2444 SIGCHLD signals. This is not a major problem, because they are used only to
2445 wake it up if nothing else does.
2447 16. A malformed maildirsize file could cause Exim to calculate negative values
2448 for the mailbox size or file count. Odd effects could occur as a result.
2449 The maildirsize information is now recalculated if the size or filecount
2452 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2453 support for a long time. Removed HAVE_SYS_VFS_H.
2455 18. Updated exipick to current release from John Jetmore.
2457 19. Allow an empty sender to be matched against a lookup in an address list.
2458 Previously the only cases considered were a regular expression, or an
2461 20. Exim went into a mad DNS lookup loop when doing a callout where the
2462 host was specified on the transport, if the DNS lookup yielded more than
2465 21. The RFC2047 encoding function was originally intended for short strings
2466 such as real names; it was not keeping to the 75-character limit for
2467 encoded words that the RFC imposes. It now respects the limit, and
2468 generates multiple encoded words if necessary. To be on the safe side, I
2469 have increased the buffer size for the ${rfc2047: expansion operator from
2472 22. Failure to deliver a bounce message always caused it to be frozen, even if
2473 there was an errors_to setting on the router. The errors_to setting is now
2476 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2477 canonical form (fully expanded) before being placed in
2478 $sender_host_address.
2480 24. Updated eximstats to version 1.33
2482 25. Include certificate and key file names in error message when GnuTLS fails
2483 to set them up, because the GnuTLS error message doesn't include the name
2484 of the failing file when there is a problem reading it.
2486 26. Expand error message when OpenSSL has problems setting up cert/key files.
2489 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2490 (this can affect the format of dates).
2492 28. exim_tidydb, when checking for the continued existence of a message for
2493 which it has found a message-specific retry record, was not finding
2494 messages that were in split spool directories. Consequently, it was
2495 deleting retry records that should have stayed in existence.
2497 29. eximstats updated to version 1.35
2498 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2499 1.35 - bugfix such that pie charts by volume are generated correctly
2501 30. The SPA authentication driver was not abandoning authentication and moving
2502 on to the next authenticator when an expansion was forced to fail,
2503 contradicting the general specification for all authenticators. Instead it
2504 was generating a temporary error. It now behaves as specified.
2506 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2507 (the order specifies the preference for clients). The order is now AES256,
2508 AES128, 3DES, ARCFOUR128.
2510 31. Small patch to Sieve code - explicitly set From: when generating an
2513 32. Exim crashed if a remote delivery caused a very long error message to be
2514 recorded - for instance if somebody sent an entire SpamAssassin report back
2515 as a large number of 550 error lines. This bug was coincidentally fixed by
2516 increasing the size of one of Exim's internal buffers (big_buffer) that
2517 happened as part of the Exiscan merge. However, to be on the safe side, I
2518 have made the code more robust (and fixed the comments that describe what
2521 33. Some experimental protocols are using DNS PTR records for new purposes. The
2522 keys for these records are domain names, not reversed IP addresses. The
2523 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2524 leaves it alone. Component reversal etc. now happens only for IP addresses.
2527 34. The host_aton() function is supposed to be passed a string that is known
2528 to be a valid IP address. However, in the case of IPv6 addresses, it was
2529 not checking this. This is a hostage to fortune. Exim now panics and dies
2530 if the condition is not met. A case was found where this could be provoked
2531 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2532 components; fortuitously, this particular loophole had already been fixed
2533 by change 4.50/55 or 4.44/33 above.
2535 If there are any other similar loopholes, the new check in host_aton()
2536 itself should stop them being exploited. The report I received stated that
2537 data on the command line could provoke the exploit when Exim was running as
2538 exim, but did not say which command line option was involved. All I could
2539 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2540 running as the user.
2543 35. There was a buffer overflow vulnerability in the SPA authentication code
2544 (which came originally from the Samba project). I have added a test to the
2545 spa_base64_to_bits() function which I hope fixes it.
2548 36. The daemon start-up calls getloadavg() while still root for those OS that
2549 need the first call to be done as root, but it missed one case: when
2550 deliver_queue_load_max is set with deliver_drop_privilege. This is
2551 necessary for the benefit of the queue runner, because there is no re-exec
2552 when deliver_drop_privilege is set.
2554 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2555 was in use, was not putting the data itself into the right store pool;
2556 consequently, it could be overwritten for a subsequent message in the same
2557 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2560 38. Sometimes the final signoff response after QUIT could fail to get
2561 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2562 < 0 before doing a fflush(). This bug looks as though it goes back to the
2563 introduction of TLS in release 3.20, but "sometimes" must have been rare
2564 because the tests only now provoked it.
2570 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2571 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2572 Now it returns an integer. A number of calls were still expecting a T/F
2573 return. Fortuitously, in all cases, the tests worked in OK situations,
2574 which is the norm. However, things would have gone wrong on any write
2575 failures on the smtp file descriptor. This function is used when sending
2576 messages over SMTP and also when doing verify callouts.
2578 2. When Exim is called to do synchronous delivery of a locally submitted
2579 message (the -odf or -odi options), it no longer closes stderr before doing
2582 3. Implemented the mua_wrapper option.
2584 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2586 5. Implemented the functions header_remove(), header_testname(),
2587 header_add_at_position(), and receive_remove_recipient(), and exported them
2590 6. If an ACL "warn" statement specified the addition of headers, Exim already
2591 inserted X-ACL-Warn: at the start if there was no header name. However, it
2592 was not making this test for the second and subsequent header lines if
2593 there were newlines in the string. This meant that an invalid header could
2594 be inserted if Exim was badly configured.
2596 7. Allow an ACL "warn" statement to add header lines at the start or after all
2597 the Received: headers, as well as at the end.
2599 8. Added the rcpt_4xx retry error code.
2601 9. Added postmaster_mailfrom=xxx to callout verification option.
2603 10. Added mailfrom=xxxx to the callout verification option, for verify=
2606 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2607 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2609 12. Inserted some casts to stop certain compilers warning when using pointer
2610 differences as field lengths or precisions in printf-type calls (mostly
2611 affecting debugging statements).
2613 13. Added optional readline() support for -be (dynamically loaded).
2615 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2616 same clock tick as a message's arrival, so that its received time was the
2617 same as the "first fail" time on the retry record, and that message
2618 remained on the queue past the ultimate address timeout, every queue runner
2619 would try a delivery (because it was past the ultimate address timeout) but
2620 after another failure, the ultimate address timeout, which should have then
2621 bounced the address, did not kick in. This was a "< instead of <=" error;
2622 in most cases the first failure would have been in the next clock tick
2623 after the received time, and all would be well.
2625 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2626 being recognized when the domain list was tested by the match_domain
2627 condition in an expansion string.
2629 16. Added the ${str2b64: operator.
2631 17. Exim was always calling setrlimit() to set a large limit for the number of
2632 processes, without checking whether the existing limit was already
2633 adequate. (It did check for the limit on file descriptors.) Furthermore,
2634 errors from getrlimit() and setrlimit() were being ignored. Now they are
2635 logged to the main and panic logs, but Exim does carry on, to try to do its
2636 job under whatever limits there are.
2638 18. Imported PCRE 5.0.
2640 19. Trivial typo in log message " temporarily refused connection" (the leading
2643 20. If the log selector return_path_on_delivery was set and an address was
2644 redirected to /dev/null, the delivery process crashed because it assumed
2645 that a return path would always be set for a "successful" delivery. In this
2646 case, the whole delivery is bypassed as an optimization, and therefore no
2649 21. Internal re-arrangement: the function for sending a challenge and reading
2650 a response while authentication was assuming a zero-terminated challenge
2651 string. It's now changed to take a pointer and a length, to allow for
2652 binary data in such strings.
2654 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2656 23. Exim was not respecting finduser_retries when seeking the login of the
2657 uid under which it was called; it was always trying 10 times. (The default
2658 setting of finduser_retries is zero.) Also, it was sleeping after the final
2659 failure, which is pointless.
2661 24. Implemented tls_on_connect_ports.
2663 25. Implemented acl_smtp_predata.
2665 26. If the domain in control=submission is set empty, Exim assumes that the
2666 authenticated id is a complete email address when it generates From: or
2667 Sender: header lines.
2669 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2670 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2671 chown and chgrp in /bin and hostname in /usr/bin.
2673 28. Exim was keeping the "process log" file open after each use, just as it
2674 does for the main log. This opens the possibility of it remaining open for
2675 long periods when the USR1 signal hits a daemon. Occasional processlog
2676 errors were reported, that could have been caused by this. Anyway, it seems
2677 much more sensible not to leave this file open at all, so that is what now
2680 29. The long-running daemon process does not normally write to the log once it
2681 has entered its main loop, and it closes the log before doing so. This is
2682 so that log files can straightforwardly be renamed and moved. However,
2683 there are a couple of unusual error situations where the daemon does write
2684 log entries, and I had neglected to close the log afterwards.
2686 30. The text of an SMTP error response that was received during a remote
2687 delivery was being truncated at 512 bytes. This is too short for some of
2688 the long messages that one sometimes sees. I've increased the limit to
2691 31. It is now possible to make retry rules that apply only when a message has a
2692 specific sender, in particular, an empty sender.
2694 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2695 it possible to be selective about when SMTP synchronization is enforced.
2697 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2699 32. Implemented hosts_connection_nolog.
2701 33. Added an ACL for QUIT.
2703 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2706 35. Added mailbox_size and mailbox_filecount to appendfile.
2708 36. Added control = no_multiline_responses to ACLs.
2710 37. There was a bug in the logic of the code that waits for the clock to tick
2711 in the case where the clock went backwards by a substantial amount such
2712 that the microsecond fraction of "now" was more than the microsecond
2713 fraction of "then" (but the whole seconds number was less).
2715 38. Added support for the libradius Radius client library this is found on
2716 FreeBSD (previously only the radiusclient library was supported).
2722 1. When certain lookups returned multiple values in the form name=value, the
2723 quoting of the values was not always being done properly. Specifically:
2724 (a) If the value started with a double quote, but contained no whitespace,
2726 (b) If the value contained whitespace other than a space character (i.e.
2727 tabs or newlines or carriage returns) it was not quoted.
2728 This fix has been applied to the mysql and pgsql lookups by writing a
2729 separate quoting function and calling it from the lookup code. The fix
2730 should probably also be applied to nisplus, ibase and oracle lookups, but
2731 since I cannot test any of those, I have not disturbed their existing code.
2733 2. A hit in the callout cache for a specific address caused a log line with no
2734 reason for rejecting RCPT. Now it says "Previous (cached) callout
2735 verification failure".
2737 3. There was an off-by-one bug in the queryprogram router. An over-long
2738 return line was truncated at 256 instead of 255 characters, thereby
2739 overflowing its buffer with the terminating zero. As well as fixing this, I
2740 have increased the buffer size to 1024 (and made a note to document this).
2742 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2743 when Exim is waiting for an SMTP response from a remote server, Exim
2744 restarts its select() call on the socket, thereby resetting its timeout.
2745 This is not a problem when such interrupts are rare. Somebody set up a cron
2746 job to run exiwhat every 2 minutes, which is less than the normal select()
2747 timeout (5 or 10 minutes). This meant that the select() timeout never
2748 kicked in because it was always reset. I have fixed this by comparing the
2749 time when an interrupt arrives with the time at the start of the first call
2750 to select(). If more time than the timeout has elapsed, the interrupt is
2751 treated as a timeout.
2753 5. Some internal re-factoring in preparation for the addition of Sieve
2754 extensions (by MH). In particular, the "personal" test is moved to a
2755 separate function, and given an option for scanning Cc: and Bcc: (which is
2756 not set for Exim filters).
2758 6. When Exim created an email address using the login of the caller as the
2759 local part (e.g. when creating a From: or Sender: header line), it was not
2760 quoting the local part when it contained special characters such as @.
2762 7. Installed new OpenBSD configuration files.
2764 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2765 try to make them clearer.
2767 9. Callout options, other than the timeout value, were being ignored when
2768 verifying sender addresses in header lines. For example, when using
2770 verify = header_sender/callout=no_cache
2772 the cache was (incorrectly) being used.
2774 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2775 only the Cygwin environment.
2777 11. When return_path_on_delivery was set as a log selector, if different remote
2778 addresses in the same message used different return paths and parallel
2779 remote delivery occurred, the wrong values would sometimes be logged.
2780 (Whenever a remote delivery process finished, the return path value from
2781 the most recently started remote delivery process was logged.)
2783 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2784 lines when AUTH and/or TLS are in use. This is the "received protocol"
2785 field. Exim used to use "asmtp" for authenticated SMTP, without any
2786 indication (in the protocol name) for TLS use. Now it follows the RFC and
2787 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2788 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2789 names appear in log lines as well as in Received: header lines.
2791 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2792 extensions, and comparison tests, and to fix some bugs.
2794 14. Changes to the "personal" filter test:
2796 (1) The test was buggy in that it was just doing the equivalent of
2797 "contains" tests on header lines. For example, if a user's address was
2798 anne@some.where, the "personal" test would incorrectly be true for
2800 To: susanne@some.where
2802 This test is now done by extracting each address from the header in turn,
2803 and checking the entire address. Other tests that are part of "personal"
2804 are now done using regular expressions (for example, to check local parts
2805 of addresses in From: header lines).
2807 (2) The list of non-personal local parts in From: addresses has been
2808 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2809 taken from the Sieve specification recommendations.
2811 (3) If the message contains any header line starting with "List-" it is
2812 treated as non-personal.
2814 (4) The test for "circular" in the Subject: header line has been removed
2815 because it now seems ill-conceived.
2817 15. Minor typos in src/EDITME comments corrected.
2819 16. Installed latest exipick from John Jetmore.
2821 17. If headers_add on a router specified a text string that was too long for
2822 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2823 of string_sprintf() is now avoided.
2825 18. $message_body_size was not set (it was always zero) when running the DATA
2826 ACL and the local_scan() function.
2828 19. For the "mail" command in an Exim filter, no default was being set for
2829 the once_repeat time, causing a random time value to be used if "once" was
2830 specified. (If the value happened to be <= 0, no repeat happened.) The
2831 default is now 0s, meaning "never repeat". The "vacation" command was OK
2832 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2833 (I found it when inspecting the code).
2835 20. There is now an overall timeout for performing a callout verification. It
2836 defaults to 4 times the callout timeout, which applies to individual SMTP
2837 commands during the callout. The overall timeout applies when there is more
2838 than one host that can be tried. The timeout is checked before trying the
2839 next host. This prevents very long delays if there are a large number of
2840 hosts and all are timing out (e.g. when the network connections are timing
2841 out). The value of the overall timeout can be changed by specifying an
2842 additional sub-option for "callout", called "maxwait". For example:
2844 verify = sender/callout=5s,maxwait=20s
2846 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2847 to the end before writing, but this should make it even safer).
2849 22. Exim was forgetting that it had advertised PIPELINING for the second and
2850 subsequent messages on an SMTP connection. It was also not resetting its
2851 memory on STARTTLS and an internal HELO.
2853 23. When Exim logs an SMTP synchronization error within a session, it now
2854 records whether PIPELINING has been advertised or not.
2856 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2857 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2858 rather than long int.
2860 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2862 26. Added the never_mail option to autoreply.
2868 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2869 crash if the getsockname() call failed; this can happen if a connection is
2870 closed very soon after it is established. The problem was simply in the
2871 order in which certain operations were done, causing Exim to try to write
2872 to the SMTP stream before it had set up the file descriptor. The bug has
2873 been fixed by making things happen in the correct order.
2879 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2880 before the connection was closed, thus losing the rejection response.
2882 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2883 some early Solaris releases, but causes trouble in current releases where
2884 socklen_t is defined.
2886 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2889 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2892 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2893 that releasing the top of it at the end releases what was used for sub-
2894 expansions (unless the block got too big). However, discard this block if
2895 the first thing is a variable or header, so that we can use its block when
2896 it is dynamic (useful for very large $message_headers, for example).
2898 6. Lookups now cache *every* query, not just the most recent. A new, separate
2899 store pool is used for this. It can be recovered when all lookup caches are
2900 flushed. Lookups now release memory at the end of their result strings.
2901 This has involved some general refactoring of the lookup sources.
2903 7. Some code has been added to the store_xxx() functions to reduce the amount
2904 of flapping under certain conditions.
2906 8. log_incoming_interface used to affect only the <= reception log lines. Now
2907 it causes the local interface and port to be added to several more SMTP log
2908 lines, for example "SMTP connection from", and rejection lines.
2910 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2912 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2914 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2915 could be overwritten at the end of the current message (or the start of a
2916 new message if it was set in a HELO ACL). The value is now preserved for
2917 the duration of the SMTP connection.
2919 12. If a transport had a headers_rewrite setting, and a matching header line
2920 contained an unqualified address, that address was qualified, even if it
2921 did not match any rewriting rules. The underlying bug was that the values
2922 of the flags that permit the existence of unqualified sender and recipient
2923 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2924 non-local messages, and by -bnq for local messages) were not being
2925 preserved with the message after it was received.
2927 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2928 "next input=" as part of the text comprising the host identity instead of
2929 the correct text. The code was using the same buffer for two different
2930 strings. However, depending on which order the printing function evaluated
2931 its arguments, the bug did not always show up. Under Linux, for example, my
2932 test suite worked just fine.
2934 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2935 doesn't work with some older versions of Perl. It has been changed to "my",
2936 which in any case is probably the better facility to use.
2938 15. A really picky compiler found some instances of statements for creating
2939 error messages that either had too many or two few arguments for the format
2942 16. The size of the buffer for calls to the DNS resolver has been increased
2943 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2944 for addresses that have a lot of PTR records. This alleviates a problem; it
2945 does not fully solve it.
2947 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2948 buffer now truncates the list and logs the incident, which is the same
2949 action as happens when Exim is looking up a host name and its aliases.
2950 Previously in this situation something unpredictable would happen;
2951 sometimes it was "internal error: store_reset failed".
2953 18. If a server dropped the connection unexpectedly when an Exim client was
2954 using GnuTLS and trying to read a response, the client delivery process
2955 crashed while trying to generate an error log message.
2957 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2958 string, for example:
2960 warn message = H1: something\nH2: something
2962 the text was added as a single header line from Exim's point of view
2963 though it ended up OK in the delivered message. However, searching for the
2964 second and subsequent header lines using $h_h2: did not work. This has been
2965 fixed. Similarly, if a system filter added multiple headers in this way,
2966 the routers could not see them.
2968 20. Expanded the error message when iplsearch is called with an invalid key to
2969 suggest using net-iplsearch in a host list.
2971 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2972 ACLs are no longer actually imposed (and a message to that effect is
2975 22. If a "gecos" field in a passwd entry contained escaped characters, in
2976 particular, if it contained a \" sequence, Exim got it wrong when building
2977 a From: or a Sender: header from that name. A second bug also caused
2978 incorrect handling when an unquoted " was present following a character
2979 that needed quoting.
2981 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2982 was not being matched caselessly.
2984 24. Arranged for all hyphens in the exim.8 source to be escaped with
2987 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2988 properly. Recipient callout cache records were still being keyed to include
2989 the sender, even when use_sender was set false. This led to far more
2990 callouts that were necessary. The sender is no longer included in the key
2991 when use_sender is false.
2993 26. Added "control = submission" modifier to ACLs.
2995 27. Added the ${base62d: operator to decode base 62 numbers.
2997 28. dnsdb lookups can now access SRV records.
2999 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3000 the configuration file.
3002 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3003 (-v) mode. This makes the output for a verbose queue run more intelligible.
3005 31. Added a use_postmaster feature to recipient callouts.
3007 32. Added the $body_zerocount variable, containing the number of binary zero
3008 bytes in the message body.
3010 33. The time of last modification of the "new" subdirectory is now used as the
3011 "mailbox time last read" when there is a quota error for a maildir
3014 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3016 35. Added +ignore_unknown as a special item in host lists.
3018 36. Code for decoding IPv6 addresses in host lists is now included, even if
3019 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3020 address was recognized as an IP address, but was then not correctly decoded
3021 into binary, causing unexpected and incorrect effects when compared with
3028 1. Very minor rewording of debugging text in manualroute to say "list of
3029 hosts" instead of "hostlist".
3031 2. If verify=header_syntax was set, and a header line with an unqualified
3032 address (no domain) and a large number of spaces between the end of the
3033 name and the colon was received, the reception process suffered a buffer
3034 overflow, and (when I tested it) crashed. This was caused by some obsolete
3035 code that should have been removed. The fix is to remove it!
3037 3. When running in the test harness, delay a bit after writing a bounce
3038 message to get a bit more predictability in the log output.
3040 4. Added a call to search_tidyup() just before forking a reception process. In
3041 theory, someone could use a lookup in the expansion of smtp_accept_max_
3042 per_host which, without the tidyup, could leave open a database connection.
3044 5. Added the variables $recipient_data and $sender_data which get set from a
3045 lookup success in an ACL "recipients" or "senders" condition, or a router
3046 "senders" option, similar to $domain_data and $local_part_data.
3048 6. Moved the writing of debug_print from before to after the "senders" test
3051 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3052 problems for message scanning, either using a data ACL, or using
3053 local_scan() because the Received: header was not generated till after they
3054 were called (in order to set the time as the time of reception completion).
3055 I have revised the way this works. The header is now generated after the
3056 body is received, but before the ACL or local_scan() are called. After they
3057 are run, the timestamp in the header is updated.
3063 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3064 before starting a queue runner without re-exec. This happened only when
3065 deliver_drop_privilege was set or when the Exim user was set to root. The
3066 effect of the bug was that timeouts during subsequent deliveries caused
3067 crashes instead of being properly handled. The handler is now left at its
3068 default (and expected) setting.
3070 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3071 message, again when deliver_drop_privilege is set or Exim is run as root.
3072 The bug described in (1) was not present in this case, but the tidying up
3073 of the other signals was missing. I have made the two cases consistent.
3075 3. The ignore_target_hosts setting on a manualroute router was being ignored
3076 for hosts that were looked up using the /MX notation.
3078 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3081 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3082 operated on the sender address. After changing the $sender_address to <>
3083 for the sender address verify, Exim was re-instated it as the original
3084 (before rewriting) address, but remembering that it had rewritten it, so it
3085 wasn't rewriting it again. This bug also had the effect of breaking the
3086 sender address verification caching when the sender address was rewritten.
3088 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3089 This has been changed so that if the ip literal address matches
3090 ignore_target_hosts, the router declines.
3092 7. Added expansion conditions match_domain, match_address, and match_local_
3093 part (NOT match_host).
3095 8. The placeholder for the Received: header didn't have a length field set.
3097 9. Added code to Exim itself and to exim_lock to test for a specific race
3098 condition that could lead to file corruption when using MBX delivery. The
3099 issue is with the lockfile that is created in /tmp. If this file is removed
3100 after a process has opened it but before that process has acquired a lock,
3101 there is the potential for a second process to recreate the file and also
3102 acquire a lock. This could lead to two Exim processes writing to the file
3103 at the same time. The added code performs the same test as UW imapd; it
3104 checks after acquiring the lock that its file descriptor still refers to
3105 the same named file.
3107 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3108 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3111 11. Added the smtp_active_hostname option. If used, this will typically be made
3112 to depend on the incoming interface address. Because $interface_address is
3113 not set up until the daemon has forked a reception process, error responses
3114 that can happen earlier (such as "too many connections") no longer contain
3117 12. If an expansion in a condition on a "warn" statement fails because a lookup
3118 defers, the "warn" statement is abandoned, and the next ACL statement is
3119 processed. Previously this caused the whole ACL to be aborted.
3121 13. Added the iplsearch lookup type.
3123 14. Added ident_timeout as a log selector.
3125 15. Added tls_certificate_verified as a log selector.
3127 16. Added a global option tls_require_ciphers (compare the smtp transport
3128 option of the same name). This controls incoming TLS connections.
3130 17. I finally figured out how to make tls_require_ciphers do a similar thing
3131 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3132 before starting the TLS session.
3134 18. Tabs are now shown as \t in -bP output.
3136 19. If the log selector return_path_on_delivery was set, Exim crashed when
3137 bouncing a message because it had too many Received: header lines.
3139 20. If two routers both had headers_remove settings, and the first one included
3140 a superfluous trailing colon, the final name in the first list and the
3141 first name in the second list were incorrectly joined into one item (with a
3142 colon in the middle).
3148 1. Added -C and -D options to the exinext utility, mainly to make it easier
3149 to include in the automated testing, but these could be helpful when
3150 multiple configurations are in use.
3152 2. The exinext utility was not formatting the output nicely when there was
3153 an alternate port involved in the retry record key, nor when there was a
3154 message id as well (for retries that were specific to a specific message
3155 and a specific host). It was also confused by IPv6 addresses, because of
3156 the additional colons they contain. I have fixed the IPv4 problem, and
3157 patched it up to do a reasonable job for IPv6.
3159 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3160 delivery, the log line now contains "pipelined" if PIPELINING was used.
3162 4. An SMTP transport process used to panic and die if the bind() call to set
3163 an explicit outgoing interface failed. This has been changed; it is now
3164 treated in the same way as a connect() failure.
3166 5. A reference to $sender_host_name in the part of a conditional expansion
3167 that was being skipped was still causing a DNS lookup. This no longer
3170 6. The def: expansion condition was not recognizing references to header lines
3171 that used bh_ and bheader_.
3173 7. Added the _cache feature to named lists.
3175 8. The code for checking quota_filecount in the appendfile transport was
3176 allowing one more file than it should have been.
3178 9. For compatibility with Sendmail, the command line option
3186 and sets the incoming protocol and host name (for trusted callers). The
3187 host name and its colon can be omitted when only the protocol is to be set.
3188 Note the Exim already has two private options, -pd and -ps, that refer to
3189 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3190 "s", but I don't think that's a major issue.
3192 10. A number of refactoring changes to the code, none of which should affect
3195 (a) The number of logging options was getting close to filling up the
3196 32-bit word that was used as a bit map. I have split them into two classes:
3197 those that are passed in the argument to log_write(), and those that are
3198 only ever tested independently outside of that function. These are now in
3199 separate 32-bit words, so there is plenty of room for expansion again.
3200 There is no change in the user interface or the logging behaviour.
3202 (b) When building, for example, log lines, the code previously used a
3203 macro that called string_cat() twice, in order to add two strings. This is
3204 not really sufficiently general. Furthermore, there was one instance where
3205 it was actually wrong because one of the argument was used twice, and in
3206 one call a function was used. (As it happened, calling the function twice
3207 did not affect the overall behaviour.) The macro has been replaced by a
3208 function that can join an arbitrary number of extra strings onto a growing
3211 (c) The code for expansion conditions now uses a table and a binary chop
3212 instead of a serial search (which was left over from when there were very
3213 few conditions). Also, it now recognizes conditions like "pam" even when
3214 the relevant support is not compiled in: a suitably worded error message is
3215 given if an attempt is made to use such a condition.
3217 11. Added ${time_interval:xxxxx}.
3219 12. A bug was causing one of the ddress fields not to be passed back correctly
3220 from remote delivery subprocesses. The field in question was not being
3221 subsequently used, so this caused to problems in practice.
3223 13. Added new log selectors queue_time and deliver_time.
3225 14. Might have fixed a bug in maildirsizefile handling that threw up
3226 "unexpected character" debug warnings, and recalculated the data
3227 unnecessarily. In any case, I expanded the warning message to give more
3230 15. Added the message "Restricted characters in address" to the statements in
3231 the default ACL that block characters like @ and % in local parts.
3233 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3234 Three changes have been made:
3236 (a) There was a serious bug; a negative response to MAIL caused the whole
3237 recipient domain to be cached as invalid, thereby blocking all messages
3238 to all local parts at the same domain, from all senders. This bug has
3239 been fixed. The domain is no longer cached after a negative response to
3240 MAIL if the sender used is not empty.
3242 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3245 (c) A new callout option, "use_sender" has been added for people who want
3246 the modified behaviour.
3252 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3255 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3256 indeed breaks things for older releases.
3258 3. Added additional logging to the case where there is a problem reading data
3259 from a filter that is running in a subprocess using a pipe, in order to
3260 try to track down a specific problem.
3262 4. Testing facility fudge: when running in the test harness and attempting
3263 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3264 getting "No route to host". Convert this to a timeout.
3266 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3269 6. Some OS don't have socklen_t but use size_t instead. This affects the
3270 fifth argument of getsockopt() amongst other things. This is now
3271 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3272 can be set for individual OS. I have set it for SunOS5, OSF1, and
3273 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3274 some earlier ones do not.
3276 7. Change 4.30/15 was not doing the test caselessly.
3278 8. The standard form for an IPv6 address literal was being rejected by address
3279 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3280 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3281 this, as well as the form without the "IPv6" on the front (but only when
3282 address literals are enabled, of course).
3284 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3286 10. Exim crashed if a message with an empty sender address specified by -f
3287 encountered a router with an errors_to setting. This could be provoked only
3288 by a command such as
3292 where an empty string was supplied; "<>" did not hit this bug.
3294 11. Installed PCRE release 4.5.
3296 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3297 remained set. It is now erased.
3299 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3300 times from message ids (which are base 36 rather than the normal 62).
3302 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3303 were being counted as actual protocol errors, and logged if the log
3304 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3305 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3306 and DATA following a set of rejected RCPTs do not count as protocol errors.
3307 In other words, Exim assumes they were pipelined, though this may not
3308 actually be the case. Of course, in all cases the client gets an
3309 appropriate error code.
3311 15. If a lookup fails in an ACL condition, a message about the failure may
3312 be available; it is used if testing the ACL cannot continue, because most
3313 such messages specify what the cause of the deferral is. However, some
3314 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3315 that caused an old message to be retained and used if a later statement
3316 caused a defer, replacing the real cause of the deferral.
3318 16. If an IP address had so many PTR records that the DNS lookup buffer
3319 was not large enough to hold them, Exim could crash while trying to process
3320 the truncated data. It now detects and logs this case.
3322 17. Further to 4.21/58, another change has been made: if (and only if) the
3323 first line of a message (the first header line) ends with CRLF, a bare LF
3324 in a subsequent header line has a space inserted after it, so as not to
3325 terminate the header.
3327 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3328 unnecessarily, used atoi() instead of strtol(), and didn't check the
3329 termination when getting file sizes from file names by regex.
3331 19. Completely re-implemented the support for maildirsize files, in the light
3332 of a number of problems with the previous contributed implementation
3333 (4.30/29). In particular:
3335 . If the quota is zero, the maildirsize file is maintained, but no quota is
3338 . If the maildir directory does not exist, it is created before any attempt
3339 to write a maildirsize file.
3341 . The quota value in the file is just a cache; if the quota is changed in
3342 the transport, the new value overrides.
3344 . A regular expression is available for excluding directories from the
3347 20. The autoreply transport checks the characters in options that define the
3348 message's headers; it allows continued headers, but it was checking with
3349 isspace() after an embedded newline instead of explicitly looking for a
3352 21. If all the "regular" hosts to which an address was routed had passed their
3353 expiry times, and had not reached their retry times, the address was
3354 bounced, even if fallback hosts were defined. Now Exim should go on to try
3357 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3358 equivalent code in the SMTP transport. Some hosts send humungous responses
3359 to HELO/EHLO, more than 1024 it seems.
3361 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3362 gives compiler warnings in some environments. I've now done it "properly",
3365 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3366 (because of problems with the built-in one) was declared to return uschar *
3367 instead of char *, causing compiler failure.
3369 25. Fixed a file descriptor leak when processing alias/forward files.
3371 26. Fixed a minor format string issue in dbfn.c.
3373 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3375 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3376 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3379 29. When a sender address is verified, it is cached, to save repeating the test
3380 when there is more than one recipient in a message. However, when the
3381 verification involves a callout, it is possible for different callout
3382 options to be set for different recipients. It is too complicated to keep
3383 track of this in the cache, so now Exim always runs a verification when a
3384 callout is required, relying on the callout cache for the optimization.
3385 The overhead is duplication of the address routing, but this should not be
3388 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3389 to be verified with callout=postmaster, and the main callout worked but the
3390 postmaster check failed, the verification correctly failed. However, if a
3391 subsequent RCPT command asked for sender verification *without* the
3392 postmaster check, incorrect caching caused this verification also to fail,
3395 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3396 it was not caching the DNS options (qualify_single, search_parents) that
3397 were used when the lookup failed. A subsequent lookup with different
3398 options therefore always gave the same answer, though there were cases
3399 where it should not have. (Example: a "domains = !$mx_any" option on a
3400 dnslookup router: the "domains" option is always processed without any
3401 widening, but the router might have qualify_single set.) Now Exim uses the
3402 cached value only when the same options are set.
3404 32. Added John Jetmore's "exipick" utility to the distribution.
3406 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3407 than a timeout (e.g. a certificate is required, and is not provided), an
3408 Exim server now closes the connection immediately. Previously it waited for
3409 the client to close - but if the client is SSL, it seems that they each
3410 wait for each other, leading to a delay before one of them times out.
3412 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3413 maintained 0.8.x compatibility because I don't think many are using it, and
3414 it is clearly obsolete.
3416 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3419 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3420 client certificate was expired. A simple patch fixes this, though I don't
3421 understand the full logic of why the verify callback is called multiple
3424 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3425 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3426 which causes problems with some clients (such as the Certicom SSL Plus
3427 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3428 disables the coutermeasure allowing Eudora to connect."
3430 38. Exim was not checking that a write() to a log file succeeded. This could
3431 lead to Bad Things if a log got too big, in particular if it hit a file
3432 size limit. Exim now panics and dies if it cannot write to a log file, just
3433 as it does if it cannot open a log file.
3435 39. Modified OS/Makefile-Linux so that it now contains
3437 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3439 The two -D definitions ensure that Exim is compiled with large file
3440 support, which makes it possible to handle log files that are bigger than
3443 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3444 instance) a domain was checked against a named list that involved a lookup,
3445 causing $domain_data to be set, then another domain was checked against the
3446 same list, then the first domain was re-checked, the value of $domain_data
3447 after the final check could be wrong. In particular, if the second check
3448 failed, it could be set empty. This bug probably also applied to
3451 41. The strip_trailing_dot option was not being applied to the address given
3452 with the -f command-line option.
3454 42. The code for reading a message's header from the spool was incrementing
3455 $received_count, but never initializing it. This meant that the value was
3456 incorrect (doubled) while delivering a message in the same process in which
3457 it was received. In the most common configuration of Exim, this never
3458 happens - a fresh exec is done - but it can happen when
3459 deliver_drop_privilege is set.
3461 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3462 it now includes up to 150 characters of the unexpected data in the log
3465 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3466 and building data strings. The size of both of these buffers was 10 000
3467 bytes - far larger than anybody would *ever* want, thought I. Needless to
3468 say, somebody hit the limit. I have increased the maximum line length to
3469 20 000 and the maximum data length of concatenated lines to 100 000. I have
3470 also fixed two bugs, because there was no checking on these buffers. Tsk,
3471 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3472 buffer is too small.
3474 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3475 lsearch lookups. Now it does.
3477 46. When parsing a route_list item in a manualroute router, a fixed-length
3478 buffer was used for the list of hosts. I made this 1024 bytes long,
3479 thinking that nobody would ever have a list of hosts that long. Wrong.
3480 Somebody had a whole pile of complicated expansion conditions, and the
3481 string was silently truncated, leading to an expansion error. It turns out
3482 that it is easier to change to an unlimited length (owing to other changes
3483 that have happened since this code was originally written) than to build
3484 structure for giving a limitation error. The length of the item that
3485 expands into the list of hosts is now unlimited.
3487 47. The lsearch lookup could not handle data where the length of text line was
3488 more than 4095 characters. Such lines were truncated, leading to shortened
3489 data being returned. It should now handle lines of any length.
3491 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3492 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3495 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3496 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3497 no longer happens in comment lines. A list of replacements is now placed
3498 at the head of all of the source files, except those whose only change is
3499 to replace PERL_COMMAND in the very first #! line.
3501 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3502 messages on the queue, with a bottom-up merge sort, using code contributed
3503 by Michael Haardt. This should make operations like -bp somewhat faster on
3504 large queues. It won't affect queue runners, except when queue_run_in_order
3507 51. Installed eximstats 1.31 in the distribution.
3509 52. Added support for SRV lookups to the dnslookup router.
3511 53. If an ACL referred to $message_body or $message_body_end, the value was not
3512 reset for any messages that followed in the same SMTP session.
3514 54. The store-handling optimization for building very long strings was not
3515 differentiating between the different store pools. I don't think this
3516 actually made any difference in practice, but I've tidied it.
3518 55. While running the routers to verify a sender address, $sender_address
3519 was still set to the sender address. This is wrong, because when routing to
3520 send a bounce to the sender, it would be empty. Therefore, I have changed
3521 it so that, while verifying a sender address, $sender_address is set to <>.
3522 (There is no change to what happens when verifying a recipient address.)
3524 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3525 target A or AAAA records (if not already returned) without resetting the
3526 qualify_single or search_parents options of the DNS resolver. These are
3527 inappropriate in this case because the targets of MX and SRV records must
3528 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3529 target that, when qualified, matched something in the local domain. These
3530 two options are now turned off when doing these lookups.
3532 57. It seems that at least some releases of Reiserfs (which does not have the
3533 concept of a fixed number of inodes) returns zero and not -1 for the
3534 number of available inodes. This interacted badly with check_spool_inodes,
3535 which assumed that -1 was the "no such thing" setting. What I have done is
3536 to check that the total number of inodes is greater than zero before doing
3537 the test of how many are available.
3539 58. When a "warn" ACL statement has a log_message modifier, the message is
3540 remembered, and not repeated. This is to avoid a lot of repetition when a
3541 message has many recipients that cause the same warning to be written.
3542 Howewer, Exim was preserving the list of already written lines for an
3543 entire SMTP session, which doesn't seem right. The memory is now reset if a
3544 new message is started.
3546 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3547 debugging output unless log_rewrite was also set.
3549 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3550 of (int)(handle) when we know that handle contains (void *)(-1).
3552 61. The Exim daemon panic-logs an error return when it closes the incoming
3553 connection. However "connection reset by peer" seems to be common, and
3554 isn't really an error worthy of noting specially, so that particular error
3557 62. When Exim is trying to find all the local interfaces, it used to panic and
3558 die if the ioctl to get the interface flags failed. However, it seems that
3559 on at least one OS (Solaris 9) it is possible to have an interface that is
3560 included in the list of interfaces, but for which you get a failure error
3561 for this call. This happens when the interface is not "plumbed" into a
3562 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3563 failure of the "get flags" call assumes that the interface is down.
3565 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3566 makes life easier for people who are doing arithmetic on fields extracted
3567 from dates, where you often get leading zeros that should not be
3568 interpreted as octal.
3570 64. Added qualify_domain to the redirect router, to override the global
3573 65. If a pathologically long header line contained very many addresses (the
3574 report of this problem mentioned 10 000) and each of them was rewritten,
3575 Exim could use up a very large amount of memory. (It kept on making new
3576 copies of the header line as it rewrote, and never released the old ones.)
3577 At the expense of a bit more processing, the header rewriting function has
3578 been changed so that it no longer eats memory in this way.
3580 66. The generation of the Received: header has been moved from the time that a
3581 message starts to be received, to the time that it finishes. The timestamp
3582 in the Received: header should now be very close to that of the <= log
3583 line. There are two side-effects of this change:
3585 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3586 logged header lines no longer include the local Received: line, because
3587 it has not yet been created. The same applies to a copy of the message
3588 that is returned to a non-SMTP sender when a message is rejected.
3590 (b) When a filter file is tested using -bf, no additional Received: header
3591 is added to the test message. After some thought, I decided that this
3594 This change does not affect the value of $received_for. It is still set
3595 after address rewriting, but before local_scan() is called.
3597 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3599 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3600 gave an unhelpful panic error message, and a defer error. I have managed to
3601 change this behaviour so that it now rejects any supplied certificate,
3602 which seems right, as the list of acceptable certificates is empty.
3604 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3605 gave an unhelpful defer error. I have not managed to make this reject any
3606 supplied certificates, but the error message it gives is "no certificate
3607 supplied", which is not helpful.
3609 70. exigrep's output now also includes lines that are not associated with any
3610 message, but which match the given pattern. Implemented by a patch from
3611 Martin Sluka, which also tidied up the Perl a bit.
3613 71. Recipient callout verification, like sender verification, was using <> in
3614 the MAIL FROM command. This isn't really the right thing, since the actual
3615 sender may affect whether the remote host accepts the recipient or not. I
3616 have changed it to use the actual sender in the callout; this means that
3617 the cache record is now keyed on a recipient/sender pair, not just the
3618 recipient address. There doesn't seem to be a real danger of callout loops,
3619 since a callout by the remote host to check the sender would use <>.
3620 [SEE ABOVE: changed after hitting problems.]
3622 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3623 temporary errors. However, in the case of such a code being given after
3624 the end of a data transmission (i.e. after ".") Exim was failing to write
3625 a retry record for the message. (Yes, there was some broken host that was
3626 actually sending 8xx at this point.)
3628 73. An unknown lookup type in a host list could cause Exim to panic-die when
3629 the list was checked. (An example that provoked this was putting <; in the
3630 middle of a list instead of at the start.) If this happened during a DATA
3631 ACL check, a -D file could be left lying around. This kind of configuration
3632 error no longer causes Exim to die; instead it causes a defer errror. The
3633 incident is still logged to the main and panic logs.
3635 74. Buglet left over from Exim 3 conversion. The message "too many messages
3636 in one connection" was written to the rejectlog but not the mainlog, except
3637 when address rewriting (yes!) was being logged.
3639 75. Added write_rejectlog option.
3641 76. When a system filter was run not as root (that is, when system_filter_user
3642 was set), the values of the $n variables were not being returned to the
3643 main process; thus, they were not subsequently available in the $sn
3646 77. Added +return_path_on_delivery log selector.
3648 78. A connection timeout was being treated differently from recipients deferred
3649 when testing hosts_max_try with a message that was older than the host's
3650 retry timeout. (The host should not be counted, thus allowing all hosts to
3651 be tried at least once before bouncing.) This may have been the cause of an
3652 occasionally reported bug whereby a message would remain on the queue
3653 longer than the retry timeout, but would be bounced if a delivery was
3654 forced. I say "may" because I never totally pinned down the problem;
3655 setting up timeout/retry tests is difficult. See also the next item.
3657 79. The ultimate address timeout was not being applied to errors that involved
3658 a combination of host plus message (for example, a timeout on a MAIL
3659 command). When an address resolved to a number of possible hosts, and they
3660 were not all tried for each delivery (e.g. because of hosts_max_try), a
3661 message could remain on the queue longer than the retry timeout.
3663 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3666 81. Fixed an obscure SMTP outgoing bug which required at least the following
3667 conditions: (a) there was another message waiting for the same server;
3668 (b) the server returned 5xx to all RCPT commands in the first message so
3669 that the message was not completed; (c) the server dropped the connection
3670 or gave a negative response to the RSET that Exim sends to abort the
3671 transaction. The observed case was a dropped connection after DATA that had
3672 been sent in pipelining mode. That is, the server had advertised PIPELINING
3673 but was not implementing it correctly. The effect of the bug was incorrect
3674 behaviour, such as trying another host, and this could lead to a crash.
3680 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3681 and daemon.c were passed as pointers to ints; they should have been
3682 pointers to socklen_t variables (which are typically unsigned ints).
3684 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3687 3. Fixed a really odd bug that affected only the testing scheme; patching a
3688 certain fixed string in the binary changed the value of another string that
3689 happened to be identical to the end of the original first string.
3691 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3692 name", it returns that address as the IP address. On some operating
3693 systems (e.g. Solaris), it also passes back the IP address string as the
3694 "host name". However, on others (e.g. Linux), it passes back an empty
3695 string. Exim wasn't checking for this, and was changing the host name to an
3696 empty string, assuming it had been canonicized.
3698 5. Although rare, it is permitted to have more than one PTR record for a given
3699 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3700 all the names associated with an address, because they do in Solaris.
3701 However, it seems that they do not in Linux for data that comes from the
3702 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3703 I found this out when I moved to a new Linux workstation and tried to run
3704 the Exim test suite.
3706 To get round this problem I have changed the code so that it now does its
3707 own call to the DNS to look up PTR records when searching for a host name.
3708 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3709 addresses that are only in /etc/hosts are still found.
3711 This behaviour is, however, controlled by an option called host_lookup_
3712 order, which defaults to "bydns:byaddr". If people want to use the other
3713 order, or indeed, just use one or the other means of lookup, they can
3714 specify it in this variable.
3716 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3717 some operating systems, this comes back from gethostbyaddr() as an empty
3718 string, and this is what Exim used to test for. However, it seems that in
3719 other systems, "." is yielded. Exim now tests for this case too.
3721 7. The values of check_spool_space and check_log_space are now held internally
3722 as a number of kilobytes instead of an absolute number of bytes. If a
3723 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3724 kilobyte. This means that much larger values can be stored.
3726 8. Exim monitor: an attempt to get the action menu when not actually pointing
3727 at a message produces an empty menu entitled "No message selected". This
3728 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3729 no entries in it ("Shell widget menu has zero width and/or height"). So I
3730 have added a single, blank menu entry in this case.
3732 9. Added ${quote_local_part.
3734 10. MIME decoding is now applied to the contents of Subject: header lines when
3737 11. Now that a reference to $sender_host_address automatically causes a reverse
3738 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3739 host lookup before query-style lookups in lists that might use this
3740 variable. This has therefore been abolished, and the "net-" prefix is no
3741 longer necessary for query-style lookups.
3743 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3744 have been a typo for LFLAGS, so it has been changed.
3746 13. The install script calls Exim with "-C /dev/null" in order to find the
3747 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3748 to be output. Howeve, since Exim outputs its version number before the
3749 error, it didn't break the script. It just looked ugly. I fixed this by
3750 always allowing "-C /dev/null" if the caller is root.
3752 14. Ignore overlarge ACL variable number when reading spool file - insurance
3753 against a later release with more variables having written the file.
3755 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3756 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3757 this, as well as the form without the "IPv6" on the front.
3759 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3760 OS/Makefile-Darwin file.
3762 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3763 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3765 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3766 variable. However, some OS use ints for inodes. I've added cast to long int
3767 to get rid of the compiler warning.
3769 19. I had forgotten to lock out "/../" in configuration file names when
3770 ALT_CONFIG_PREFIX was set.
3772 20. Routers used for verification do not need to specify transports. However,
3773 if such a router generated a host list, and callout was configured, Exim
3774 crashed, because it could not find a port number from the (non-existent)
3775 transport. It now assumes port 25 in this circumstance.
3777 21. Added the -t option to exigrep.
3779 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3780 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3781 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3782 them is set, LOOKUP_LSEARCH is forced.
3784 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3786 24. Added sender and host information to the "rejected by local_scan()" log
3787 line; previously there was no indication of these.
3789 25. Added .include_if_exists.
3791 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3792 Linux. It turns out that not all file systems support this. Apparently some
3793 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3794 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3795 not supported on the file descriptor, is now ignored when Exim is trying to
3796 sync a directory. This applies only to Linux.
3798 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3800 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3801 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3804 29. Added support for maildirsize files from supplied patch (modified a bit).
3806 30. The use of :fail: followed by an empty string could lead Exim to respond to
3807 sender verification failures with (e.g.):
3809 550 Verification failed for <xxx>
3810 550 Sender verify failed
3812 where the first response line was missing the '-' that indicates it is not
3813 the final line of the response.
3815 31. The loop for finding the name of the user that called Exim had a hardwired
3816 limit of 10; it now uses the value of finduser_retries, which is used for
3817 all other user lookups.
3819 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3822 33. Exim was neglecting to zero errno before one call of strtol() when
3823 expanding a string and expecting an integer value. On some systems this
3824 resulted in spurious "integer overflow" errors. Also, it was casting the
3825 result into an int without checking.
3827 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3828 did not work. The code looks as if it has *never* worked, though it appears
3829 to have been documented since at least releast 1.62. I have made it work.
3831 35. The "timeout_DNS" error in retry rules, also documented since at least
3832 1.62, also never worked. As it isn't clear exactly what this means, and
3833 clearly it isn't a major issue, I have abolished the feature by treating it
3834 as "timeout", and writing a warning to the main and panic logs.
3836 36. The display of retry rules for -brt wasn't always showing the error code
3839 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3840 timeout_connect_A, timeout_connect_MX.
3842 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3843 to the empty sender.
3845 39. The daemon was not analysing the content of -oX till after it had closed
3846 stderr and disconnected from the controlling terminal. This meant that any
3847 syntax errors were only noted on the panic log, and the return code from
3848 the command was 0. By re-arranging the code a little, I've made the
3849 decoding happen first, so such errors now appear on stderr, and the return
3850 code is 1. However, the actual setting up of the sockets still happens in
3851 the disconnected process, so errors there are still only recorded on the
3854 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3855 connections (as happens on some IP stacks) was logged at start up time as
3856 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3857 it from "IPv6 and IPv4", which means that two separate sockets are being
3860 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3861 says whether AF_INET or AF_INET6 was passed as an argument.
3863 42. Exiwhat output was messed up when time zones were included in log
3866 43. Exiwhat now gives more information about the daemon's listening ports,
3867 and whether -tls-on-connect was used.
3869 44. The "port" option of the smtp transport is now expanded.
3871 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3872 silently ignored. Now an error message is written to the main and panic
3875 46. There's a new ACL modifier called "logwrite" which writes to a log file
3876 as soon as it is encountered.
3878 47. Added $local_user_uid and $local_user_gid at routing time.
3880 48. Exim crashed when trying to verify a sender address that was being
3883 49. Exim was recognizing only a space character after ".include". It now also
3884 recognizes a tab character.
3886 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3887 extracting the relevant information from the specification. The man page no
3888 longer contains scrambled data for the -d option, and I've added a section
3889 at the front about calling Exim under different names.
3891 51. Added "extra_headers" argument to the "mail" command in filter files.
3893 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3896 53. Installed eximstats 1.29.
3898 54. Added transport_filter_timeout as a generic transport option.
3900 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3901 Cc: header lines. This was required by RFC 822, but it not required by RFC
3904 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3905 incoming messages that did not have them. Now it does so only if the
3906 message originates locally, that is, if there is no associated remote host
3907 address. When Resent- header lines are present, this applies to the Resent-
3908 lines rather than the non-Resent- lines.
3910 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3911 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3913 58. Messages for configuration errors now include the name of the main
3914 configuration file - useful now that there may be more than one file in a
3915 list (.included file names were always shown).
3917 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3918 for those rare installations that do not start the daemon as root or run it
3919 setuid root. I've cut out the call to initgroups() if the daemon is not
3922 60. The Exim user and group can now be bound into the binary as text strings
3923 that are looked up at the start of Exim's processing.
3925 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3927 62. Added $mailstore_basename variable.
3929 63. Installed patch to sieve.c from Michael Haardt.
3931 64. When Exim failed to open the panic log after failing to open the main log,
3932 the original message it was trying to log was written to stderr and debug
3933 output, but if they were not available (the usual case in production), it
3934 was lost. Now it is written to syslog before the two lines that record the
3935 failures to open the logs.
3937 65. Users' Exim filters run in subprocesses under the user's uid. It is
3938 possible for a "deliver" command or an alias in a "personal" command to
3939 provoke an address rewrite. If logging of address rewriting is configured,
3940 this fails because the process is not running as root or exim. There may be
3941 a better way of dealing with this, but for the moment (because 4.30 needs
3942 to be released), I have disabled address rewrite logging when running a
3943 filter in a non-root, non-exim process.
3949 1. The buildconfig auxiliary program wasn't quoting the value set for
3950 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3951 not defined. This bug was masked in 4.22 by the effect that was fixed in
3954 2. Some messages that were rejected after a message id was allocated were
3955 shown as "incomplete" by exigrep. It no longer does this for messages that
3956 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3958 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3959 have allow_domain_literals set, the ID did not get logged in the <= line.
3960 Domain literals are now always recognized in Message-ID: header lines.
3962 4. The first argument for a ${extract expansion item is the key name or field
3963 number. Leading and trailing spaces in this item were not being ignored,
3964 causing some misleading effects.
3966 5. When deliver_drop_privilege was set, single queue runner processes started
3967 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3968 same command in the process it spins off) were not dropping privilege.
3970 6. When the daemon running as "exim" started a queue runner, it always
3971 re-executed Exim in the spun-off process. This is a waste of effort when
3972 deliver_drop_privilege is set. The new process now just calls the
3973 queue-runner function directly.
3979 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3982 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3983 ignored. Though the use of -oP was forcing the writing of a pid file, it
3984 was always written to the default place.
3986 3. If the message "no IP address found for host xxxx" is generated during
3987 incoming verification, it is now followed by identification of the incoming
3988 connection (so you can more easily find what provoked it).
3990 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3992 5. Added some features to "harden" Exim a bit more against certain attacks:
3994 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3995 be put in Local/Makefile. This is like the never_users runtime option,
3996 but it cannot be overridden. The default setting is "root".
3998 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3999 prefix string with which any file named in a -C command line option
4002 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4003 is retained for -C and -D only if the caller of Exim is root. Without
4004 it, the exim user may also use -C and -D and retain privilege.
4006 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4007 command line option is disabled.
4009 6. Macro names set by the -D option must start with an upper case letter, just
4010 like macro names defined in the configuration file.
4012 7. Added "dereference=" facility to LDAP.
4014 8. Two instances of the typo "uknown" in the source files are fixed.
4016 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4017 the Configure-Makefile script screwed up while processing it.
4019 10. Incorporated PCRE 4.4.
4021 11. The SMTP synchronization check was not operating right at the start of an
4022 SMTP session. For example, it could not catch a HELO sent before the client
4023 waited for the greeting. There is now a check for outstanding input at the
4024 point when the greeting is written. Because of the duplex, asynchronous
4025 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4026 way, but not yet received, when the check is performed.
4028 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4029 on TCP/IP sockets, because this apparently causes some broken clients to
4032 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4033 unchanged) from the Cygwin maintainer.
4035 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4036 was supported instead of testing for mbx. Effectively a typo.
4038 15. The spa authenticator server code was not checking that the input it
4039 received was valid base64.
4041 16. The debug output line for the "set" modifier in ACLs was not showing the
4042 name of the variable that was being set.
4044 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4046 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4047 lookup on its own. Something else was needed to trigger the lookup. For
4048 example, a match in host_lookup or the need for a host name in a host list.
4049 Now, if $sender_host_name is referenced and the host name has not yet been
4050 looked up, a lookup is performed. If the lookup fails, the variable remains
4051 empty, and $host_lookup_failed is set to "1".
4053 19. Added "eqi" as a case-independent comparison operator.
4055 20. The saslauthd authentication condition could segfault if neither service
4056 nor realm was specified.
4058 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4059 error message that was logged was misleading, and incoming SMTP
4060 connections were dropped. The message is now more accurate, and temporary
4061 errors are given to SMTP connections.
4063 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4064 (except RSET) with a 421 error, until QUIT is received. However, it was
4065 failing to send a response to QUIT.
4067 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4068 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4069 ACL is now run after the helo_try_verify_hosts code.
4071 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4072 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4073 case-independent, so other case variants are also recognized). Apparently
4074 some systems use these upper case variants.
4076 25. If more than two messages were waiting for the same host, and a transport
4077 filter was specified for the transport, Exim sent two messages over the
4078 same TCP/IP connection, and then failed with "socket operation on non-
4079 socket" when it tried to send the third.
4081 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4083 27. The extern definition of crypt16() in expand.c was not being excluded when
4084 the OS had its own crypt16() function.
4086 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4087 as a preferred synonym for return_size_limit, both as an option and as an
4090 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4092 30. Changed the default configuration ACL to relax the local part checking rule
4093 for addresses that are not in any local domains. For these addresses,
4094 slashes and pipe symbols are allowed within local parts, but the sequence
4095 /../ is explicitly forbidden.
4097 31. SPA server authentication was not clearing the challenge buffer before
4100 32. log_message in a "warn" ACL statement was writing to the reject log as
4101 well as to the main log, which contradicts the documentation and doesn't
4102 seem right (because no rejection is happening). So I have stopped it.
4104 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4105 However, I am unable to do any testing of this.
4107 34. Fixed an infelicity in the appendfile transport. When checking directories
4108 for a mailbox, to see if any needed to be created, it was accidentally
4109 using path names with one or more superfluous leading slashes; tracing
4110 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4112 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4113 contents are added to the log line that is written for every discarded
4114 recipient. (Previously a log_message setting was ignored.)
4116 36. The ${quote: operator now quotes the string if it is empty.
4118 37. The install script runs exim in order to find its version number. If for
4119 some reason other than non-existence or emptiness, which it checks, it
4120 could not run './exim', it was installing it with an empty version number,
4121 i.e. as "exim-". This error state is now caught, and the installation is
4124 38. An argument was missing from the function that creates an error message
4125 when Exim fails to connect to the socket for saslauthd authentication.
4126 This could cause Exim to crash, or give a corrupted message.
4128 39. Added isip, isip4, and isip6 to ${if conditions.
4130 40. The ACL variables $acl_xx are now saved with the message, and can be
4131 accessed later in routers, transports, and filters.
4133 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4134 strings in the file are not string-expanded.
4136 42. If a MAIL command specified a SIZE value that was too large to fit into an
4137 int variable, the check against message_size_limit failed. Such values are
4138 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4139 day this will have to be increased, but I don't think I want to be around
4140 when emails are that large.
4147 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4148 iconv() is not standard in FreeBSD.
4150 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4151 IPv6 enabled. The observed symptom was a segmentation fault on return from
4152 the function os_common_find_running_interfaces() in src/os.c.
4154 3. In the check_special_case() function in daemon.c I had used "errno" as an
4155 argument name, which causes warnings on some systems. This was basically a
4156 typo, since it was named "eno" in the comments!
4158 4. The code that waits for the clock to tick (at a resolution of some fraction
4159 of a second) so as to ensure message-id uniqueness was always waiting for
4160 at least one whole tick, when it could have waited for less. [This is
4161 almost certainly not relevant at current processor speeds, where it is
4162 unlikely to ever wait at all. But we try to future-proof.]
4164 5. The function that sleeps for a time interval that includes fractions of a
4165 second contained a race. It did not block SIGALRM between setting the
4166 timer, and suspending (a couple of lines later). If the interval was short
4167 and the sigsuspend() was delayed until after it had expired, the suspension
4168 never ended. On busy systems this could lead to processes getting stuck for
4171 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4172 process, before it forks any delivery processes. The open lookup caching
4173 mechanism meant that the open file or database connection was passed into
4174 the delivery process. The problem was that delivery processes always tidy
4175 up cached lookup data. This could cause a problem for the next delivery
4176 process started by the queue runner, because the external queue runner
4177 process does not know about the closure. So the next delivery process
4178 still has data in the lookup cache. In the case of a file lookup, there was
4179 no problem because closing a file descriptor in a subprocess doesn't affect
4180 the parent. However, if the lookup was caching a connection to a database,
4181 the connection was closed, and the second delivery process was likely to
4182 see errors such as "PGSQL: query failed: server closed the connection
4183 unexpectedly". The problem has been fixed by closing all cached lookups
4184 in a queue runner before running a delivery process.
4186 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4187 seem to have the "const" qualifier which it has on other OS. I've
4190 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4191 *of the same type* (client or server) with the same public name that an
4192 error should be diagnosed.
4194 9. When Exim looked up a host name for an IP address, but failed to find the
4195 original IP address when looking up the host name (a safety check), it
4196 output the message "<ip address> does not match any IP for NULL", which was
4197 confusing, to say the least. The bug was that the host name should have
4198 appeared instead of "NULL".
4200 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4201 user that is built into the binary, and the -C or -D options is used, root
4202 privilege is dropped before the configuration file is read. In addition,
4203 logging is switched to stderr instead of the normal log files. If the
4204 configuration then re-defines the Exim user, the unprivileged environment
4205 is probably not what is expected, so Exim logs a panic warning message (but
4208 However, if deliver_drop_privilege is set, the unprivileged state may well
4209 be exactly what is intended, so the warning has been cut out in that case,
4210 and Exim is allowed to try to write to its normal log files.
4216 1. smtp_return_error_details was not giving details for temporary sender
4217 or receiver verification errors.
4219 2. Diagnose a configuration error if two authenticators have the same public
4222 3. Exim used not to create the message log file for a message until the first
4223 delivery attempt. This could be confusing when incoming messages were held
4224 for policy or load reasons. The message log file is now created at the time
4225 the message is received, and an initial "Received" line is written to it.
4227 4. The automatically generated man page for command line options had a minor
4228 bug that caused no ill effects; however, a more serious problem was that
4229 the procedure for building the man page automatically didn't always
4230 operate. Consequently, release 4.20 contains an out-of-date version. This
4231 shouldn't happen again.
4233 5. When building Exim with embedded Perl support, the script that builds the
4234 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4235 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4237 6. The freeze_tell option was not being used for messages that were frozen on
4238 arrival, either by an ACL or by local_scan().
4240 7. Added the smtp_incomplete_transaction log selector.
4242 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4243 was accepting AUTH without a new EHLO.
4245 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4246 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4249 10. Logging of TCP/IP connections (when configured) now happens in the main
4250 daemon process instead of the child process, so that the TCP/IP connection
4251 count is more accurate (but it can never be perfect).
4253 11. The use of "drop" in a nested ACL was not being handled correctly in the
4254 outer ACL. Now, if condition failure induced by the nested "drop" causes
4255 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4256 or "require"), the connection is dropped.
4258 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4259 that yield "discard" can now be used with an "accept" or a "discard" verb,
4260 but an error is generated for any others (because I can't see a useful way
4261 to define what should happen).
4263 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4264 are now processed in the same way as lines in the Exim configuration file.
4265 In particular, continuation lines are supported.
4267 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4269 15. Added -ti meaning -t -i.
4271 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4272 domains, and warn by logging if others are found.
4274 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4275 returned from GIFCONF when Exim is trying to find the list of interfaces on
4276 a host. The code in os.c has been modified to copy each ifreq to an aligned
4277 structure in all cases.
4279 Also, in some cases, the returned ifreq's were being copied to a 'struct
4280 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4281 means the last couple of bytes of an IPv6 address could be chopped if the
4282 ifreq contained only a normal sockaddr (14 bytes storage).
4284 18. Named domain lists were not supported in the hosts_treat_as_local option.
4285 An entry such as +xxxx was not recognized, and was treated as a literal
4288 19. Ensure that header lines added by a DATA ACL are included in the reject log
4289 if the ACL subsequently rejects the message.
4291 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4292 MD5 (which is deprecated).
4294 21. When testing a filter file using -bf, Exim was writing a message when it
4295 took the sender from a "From " line in the message, but it was not doing so
4296 when it took $return_path from a Return-Path: header line. It now does.
4298 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4299 with a valid header line field name (a series of printing characters
4300 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4302 23. Changed "disc" in the source to "disk" to conform to the documentation and
4303 the book and for uniformity.
4305 24. Ignore Sendmail's -Ooption=value command line item.
4307 25. When execve() failed while trying to run a command in a pipe transport,
4308 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4309 could be confused with a return value of 69 from the command itself. This
4310 has been changed to 127, the value the shell returns if it is asked to run
4311 a non-existent command. The wording for the related log line suggests a
4312 non-existent command as the problem.
4314 26. If received_header_text expands to an empty string, do not add a Received:
4315 header line to the message. (Well, it adds a token one on the spool, but
4316 marks it "old" so that it doesn't get used or transmitted.)
4318 27. Installed eximstats 1.28 (addition of -nt option).
4320 28. There was no check for failure on the call to getsockname() in the daemon
4321 code. This can fail if there is a shortage of resources on the system, with
4322 ENOMEM, for example. A temporary error is now given on failure.
4324 29. Contrary to the C standard, it seems that in some environments, the
4325 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4326 program. Exim now does this explicitly; it affects the formatting of
4327 timestamps using strftime().
4329 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4330 complaints. I've now initialized all the variables, to avoid this.
4332 32. Header lines added by a system filter were not being "seen" during
4333 transport-time rewrites.
4335 33. The info_callback() function passed to OpenSSL is set up with type void
4336 (*)(SSL *, int, int), as described somewhere. However, when calling the
4337 function (actually a macro) that sets it up, the type void(*)() is
4338 expected. I've put in a cast to prevent warnings from picky compilers.
4340 34. If a DNS black list lookup found a CNAME record, but there were no A
4341 records associated with the domain it pointed at, Exim crashed.
4343 35. If a DNS black list lookup returned more than one A record, Exim ignored
4344 all but the first. It now scans all returned addresses if a particular IP
4345 value is being sought. In this situation, the contents of the
4346 $dnslist_value variable are a list of all the addresses, separated by a
4349 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4350 to accept a host name and all its aliases if the forward lookup for any of
4351 them yielded the IP address of the incoming connection. Now it accepts only
4352 those names whose forward lookup yields the correct IP address. Any other
4353 names are discarded. This closes a loophole whereby a rogue DNS
4354 administrator could create reverse DNS records to break through a
4355 wildcarded host restriction in an ACL.
4357 37. If a user filter or a system filter that ran in a subprocess used any of
4358 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4359 the wrong values were passed to the pipe command ($thisaddress had the
4360 value of $0, $0 had the value of $1, etc). This bug was introduced by
4361 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4363 38. Improved the line breaking for long SMTP error messages from ACLs.
4364 Previously, if there was no break point between 40 and 75 characters, Exim
4365 left the rest of the message alone. Two changes have been made: (a) I've
4366 reduced the minimum length to 35 characters; (b) if it can't find a break
4367 point between 35 and 75 characters, it looks ahead and uses the first one
4368 that it finds. This may give the occasional overlong line, but at least the
4369 remaining text gets split now.
4371 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4372 file descriptors might be low, and that setting 1000 would always raise it.
4373 It turns out that in some environments, the limit is already over 1000 and
4374 that lowering it causes trouble. So now Exim takes care not to decrease it.
4376 40. When delivering a message, the value of $return_path is set to $sender_
4377 address at the start of routing (routers may change the value). By an
4378 oversight, this default was not being set up when an address was tested by
4379 -bt or -bv, which affected the outcome if any router or filter referred to
4382 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4383 log only when "message" or "log_message" are set. However, if one of the
4384 conditions was an address verification, or a call to a nested ACL, the
4385 messages generated by the underlying test were being passed through. This
4386 no longer happens. The underlying message is available in $acl_verify_
4387 message for both "message" and "log_message" expansions, so it can be
4388 passed through if needed.
4390 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4391 new expansion $bh_ to give the encoded byte string without charset
4392 translation. Translation happens only if iconv() is available; HAVE_ICONV
4393 indicates this at build time. HEADERS_CHARSET gives the charset to
4394 translate to; headers_charset can change it in the configuration, and
4395 "headers charset" can change it in an individual filter file.
4397 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4398 that creates RFC 2047 encoded "words" labels them as that charset instead
4399 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4400 expansion operator; (ii) when Exim creates a From: line for a local
4401 message; (iii) when a header line is rewritten to include a "phrase" part.
4403 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4404 buggy, causing it to skip the first lines of messages whose message ID
4405 ended in 'D'. This would not have bitten before Exim release 4.14, because
4406 message IDs were unlikely to end in 'D' before then. The effect was to have
4407 incorrect size information for certain domains.
4409 45. #include "config.h" was missing at the start of the crypt16.c module. This
4410 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4413 46. If there was a timeout during a "random" callout check, Exim treated it as
4414 a failure of the random address, and carried on sending RSET and the real
4415 address. If the delay was just some slowness somewhere, the response to the
4416 original RCPT would be taken as a response to RSET and so on, causing
4417 mayhem of various kinds.
4419 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4420 when I implemented it. It didn't allow for the fact that some option values
4421 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4422 the right test for positive values.
4424 48. Domain names in DNS records are case-independent. Exim always looks them up
4425 in lower case. Some resolvers return domain names in exactly the case they
4426 appear in the zone file, that is, they may contain uppercase letters. Not
4427 all resolvers do this - some return always lower case. Exim was treating a
4428 change of case by a resolver as a change of domain, similar to a widening
4429 of a domain abbreviation. This triggered its re-routing code and so it was
4430 trying to route what was effectively the same domain again. This normally
4431 caused routing to fail (because the router wouldn't handle the domain
4432 twice). Now Exim checks for this case specially, and just changes the
4433 casing of the domain that it ultimately uses when it transmits the message
4436 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4439 50. If a filter generated a file delivery with a non-absolute name (possible if
4440 no home directory exists for the router), the forbid_file option was not
4443 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4444 the existing equality matching.
4446 52. Exim was using ints instead of ino_t variables in some places where it was
4447 dealing with inode numbers.
4449 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4450 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4451 called TMPDIR, and if it finds it is different, it changes its value.
4453 54. The smtp_printf() function is now made available to local_scan() so
4454 additional output lines can be written before returning. There is also an
4455 smtp_fflush() function to enable the detection of a dropped connection.
4456 The variables smtp_input and smtp_batched_input are exported to
4459 55. Changed the default runtime configuration: the message "Unknown user"
4460 has been removed from the ACL, and instead placed on the localuser router,
4461 using the cannot_route_message feature. This means that any verification
4462 failures that generate their own messages won't get overridden. Similarly,
4463 the "Unrouteable address" message that was in the ACL for unverifiable
4464 relay addresses has also been removed.
4466 56. Added hosts_avoid_esmtp to the smtp transport.
4468 57. The exicyclog script was not checking for the esoteric option
4469 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4470 will work only if exicyclog is run under the appropriate euid.
4472 58. Following a discussion on the list, the rules by which Exim recognises line
4473 endings on incoming messages have been changed. The -dropcr and drop_cr
4474 options are now no-ops, retained only for backwards compatibility. The
4475 following line terminators are recognized: LF CRLF CR. However, special
4476 processing applies to CR:
4478 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4479 nor a local message in the state where . is a terminator.
4481 (ii) If a bare CR is encountered in a header line, an extra space is added
4482 after the line terminator so as not to end the header. The reasoning
4483 behind this is that bare CRs in header lines are most likely either
4484 to be mistakes, or people trying to play silly games.
4486 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4487 was being incorrectly given as 18 bytes larger than it should have been.
4488 This is a VOB (very old bug).
4490 60. This may never have affected anything current, but just in case it has:
4491 When the local host is found other than at the start of a list of hosts,
4492 the local host, those with the same MX, and any that follow, are discarded.
4493 When the list in question was part of a longer list of hosts, the following
4494 hosts (not currently being processed) were also being discarded. This no
4495 longer happens. I'm not sure if this situation could ever has previously
4498 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4501 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4502 header. This is something that is recommended in a new Internet Draft, and
4503 is something that is documented as being done by Sendmail. There are two
4504 possible values. For messages generated by the autoreply transport, Exim
4507 Auto-Submitted: auto-replied
4509 whereas for all other generated messages (e.g. bounces) it adds
4511 Auto-Submitted: auto-generated
4513 63. The "personal" condition in filters now includes a test for the
4514 Auto-Submitted: header. If it contains the string "auto-" the message it
4515 not considered personal.
4517 64. Added rcpt_include_affixes as a generic transport option.
4519 65. Added queue_only_override (default true).
4521 66. Added the syslog_duplication option.
4523 67. If what should have been the first header line of a message consisted of
4524 a space followed by a colon, Exim was mis-interpreting it as a header line.
4525 It isn't of course - it is syntactically invalid and should therefore be
4526 treated as the start of the message body. The misbehaviour could have
4527 caused a number of strange effects, including loss of data in subsequent
4528 header lines, and spool format errors.
4530 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4531 client host had authenticated. This control can now be exercised by an ACL
4532 for more flexibility.
4534 69. By default, callouts do not happen when testing with -bh. There is now a
4535 variant, -bhc, which does actually run the callout code, including
4536 consulting and updating the callout cache.
4538 70. Added support for saslauthd authentication, courtesy of Alexander
4541 71. If statvfs() failed on the spool or log directories while checking their
4542 size for availability, Exim confusingly gave the error "space shortage".
4543 Furthermore, in debugging mode it crashed with a floating point exception.
4544 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4545 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4546 serious problem, Exim now writes to the main and panic logs when this
4547 happens, with details of the failure. It then refuses to accept the
4548 incoming message, giving the message "spool directory problem" or "log
4549 directory problem" with a 421 code for SMTP messages.
4551 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4552 0, 1, and 2 exist, because some OS complain for execs without them (see
4553 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4554 descriptors. However, the code omitted to check that the open succeeded,
4555 causing mysterious errors if for some reason the permissions on /dev/null
4556 got screwed. Now Exim writes a message to the main and panic logs, and
4557 bombs out if it can't open /dev/null.
4559 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4560 interact so that it is all more flexible. It is supposed to remain
4561 backwards compatible. Also added extra_local_interfaces.
4563 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4564 to bomb out with an assertion failure - to the client this appears as a
4565 connection drop. This problem occurs in the part of the code that was taken
4566 from the Samba project. Fortunately, the assertion is in a very simple
4567 function, so I have fixed this by reproducing the function inline in the
4568 one place where it is called, and arranging for authentication to fail
4569 instead of killing the process with assert().
4571 75. The SPA client code was not working when the server requested OEM rather
4572 than Unicode encoding.
4574 76. Added code to make require_files with a specific uid setting more usable in
4575 the case where statting the file as root fails - usually a non-root-mounted
4576 NFS file system. When this happens and the failure is EACCES, Exim now
4577 forks a subprocess and does the per-uid checking as the relevant uid.
4579 77. Added process_log_path.
4581 78. If log_file_path was not explicitly set, a setting of check_log_space or
4582 check_log_inodes was ignored.
4584 79. If a space check for the spool or log partitions fails, the incident is now
4585 logged. Of course, in the latter case the data may get lost...
4587 80. Added the %p formatting code to string_format() so that it can be used to
4588 print addresses in debug_print(). Adjusted all the address printing in the
4589 debugging in store.c to use %p rather than %d.
4591 81. There was a concern that a line of code in smtp_in.c could overflow a
4592 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4593 initially expressed, the concern was not well-founded, because trailing
4594 spaces are removed early. However, if the trailing spaces were followed by
4595 a NULL, they did not get removed, so the overflow was possible. Two fixes
4598 (a) I re-wrote the offending code in a cleaner fashion.
4599 (b) If an incoming SMTP command contains a NULL character, it is rejected
4602 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4603 runs initgroups(), so that if the Exim user is in any additional groups,
4604 they will be used during message reception.
4610 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob