1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.469 2007/02/06 10:00:24 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.
82 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
83 fixed by 4.65/MH/01 (is this a record?) are fixed:
85 (i) An empty string was always treated as zero by the numeric comparison
86 operators. This behaviour has been restored.
88 (ii) It is documented that the numeric comparison operators always treat
89 their arguments as decimal numbers. This was broken in that numbers
90 starting with 0 were being interpreted as octal.
92 While fixing these problems I realized that there was another issue that
93 hadn't been noticed. Values of message_size_limit (both the global option
94 and the transport option) were treated as octal if they started with 0.
95 The documentation was vague. These values are now always treated as
96 decimal, and I will make that clear in the documentation.
102 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
103 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
106 MH/01 Don't check that the operands of numeric comparison operators are
107 integers when their expansion is in "skipping" mode (fixes bug
108 introduced by 4.64-PH/07).
110 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
111 child addresses, Exim now panics and dies. Previously, because the count
112 is held in a short int, deliveries were likely to be lost. As such a
113 large number of recipients for a single message is ridiculous
114 (performance will be very, very poor), I have chosen to impose a limit
115 rather than extend the field.
121 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
122 leftover -K file (the existence of which was triggered by #402).
123 While we were at it, introduced process PID as part of the -K
124 filename. This should rule out race conditions when creating
127 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
128 processing considerably. Previous code took too long for large mails,
129 triggering a timeout which in turn triggers #401.
131 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
132 in the DK code in transports.c. sendfile() is not really portable,
133 hence the _LINUX specificness.
135 TF/01 In the add_headers option to the mail command in an Exim filter,
136 there was a bug that Exim would claim a syntax error in any
137 header after the first one which had an odd number of characters
140 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
141 callout verification, Exim cached a "reject" for the entire domain. This
142 is correct for most verifications, but it is not correct for a recipient
143 verification with use_sender or use_postmaster set, because in that case
144 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
145 case of MAIL FROM:<> rejection from other early rejections (e.g.
146 rejection of HELO). When verifying a recipient using a non-null MAIL
147 address, the cache is ignored if it shows MAIL FROM:<> rejection.
148 Whatever the result of the callout, the value of the domain cache is
149 left unchanged (for any other kind of callout, getting as far as trying
150 RCPT means that the domain itself is ok).
152 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
155 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
156 manifest itself as EPIPE rather than ECONNECT. When tidying away a
157 session, the daemon ignores ECONNECT errors and logs others; it now
158 ignores EPIPE as well.
160 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
161 (quoted-printable decoding).
163 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
164 later the small subsequent patch to fix an introduced bug.
166 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
168 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
170 PH/08 An error is now given if message_size_limit is specified negative.
172 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
173 to be given (somewhat) arbitrary names.
175 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
178 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
179 miscellaneous code fixes
181 PH/10 Added the log_reject_target ACL modifier to specify where to log
184 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
185 hostname. This is wrong, because it relates to the incoming message (and
186 probably the interface on which it is arriving) and not to the outgoing
187 callout (which could be using a different interface). This has been
188 changed to use the value of the helo_data option from the smtp transport
189 instead - this is what is used when a message is actually being sent. If
190 there is no remote transport (possible with a router that sets up host
191 addresses), $smtp_active_hostname is used.
193 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
194 tweaks were necessary in order to get it to work (see also 21 below):
195 (a) The code assumed that strncpy() returns a negative number on buffer
196 overflow, which isn't the case. Replaced with Exim's string_format()
198 (b) There were several signed/unsigned issues. I just did the minimum
199 hacking in of casts. There is scope for a larger refactoring.
200 (c) The code used strcasecmp() which is not a standard C function.
201 Replaced with Exim's strcmpic() function.
202 (d) The code set only $1; it now sets $auth1 as well.
203 (e) A simple test gave the error "authentication client didn't specify
204 service in request". It would seem that Dovecot has changed its
205 interface. Fortunately there's a specification; I followed it and
206 changed what the client sends and it appears to be working now.
208 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
211 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
212 address A is aliased to B and C, where B exists and C does not. Without
213 -v the output is "A verified" because verification stops after a
214 successful redirection if more than one address is generated. However,
215 with -v the child addresses are also verified. Exim was outputting "A
216 failed to verify" and then showing the successful verification for C,
217 with its parentage. It now outputs "B failed to verify", showing B's
218 parentage before showing the successful verification of C.
220 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
221 look up a TXT record in a specific list after matching in a combined
224 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
225 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
226 they consult the DNS. I had assumed they would set it the way they
227 wanted; and indeed my experiments on Linux seem to show that in some
228 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
229 To be on the safe side, however, I have now made the interface to
230 host_find_byname() similar to host_find_bydns(), with an argument
231 containing the DNS resolver options. The host_find_byname() function now
232 sets these options at its start, just as host_find_bydns() does. The smtp
233 transport options dns_qualify_single and dns_search_parents are passed to
234 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
235 of host_find_byname() use the default settings of RES_DEFNAMES
236 (qualify_single) but not RES_DNSRCH (search_parents).
238 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
239 spool_read_header() do less string testing, by means of a preliminary
240 switch on the second character of optional "-foo" lines. (This is
241 overdue, caused by the large number of possibilities that now exist.
242 Originally there were few.) While I was there, I also converted the
243 str(n)cmp tests so they don't re-test the leading "-" and the first
244 character, in the hope this might squeeze out yet more improvement.
246 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
247 flag allowing group syntax was set by the header_syntax check but not
248 turned off, possible causing trouble later; (2) The flag was not being
249 set at all for the header_verify test, causing "group"-style headers to
250 be rejected. I have now set it in this case, and also caused header_
251 verify to ignore an empty address taken from a group. While doing this, I
252 came across some other cases where the code for allowing group syntax
253 while scanning a header line wasn't quite right (mostly, not resetting
254 the flag correctly in the right place). These bugs could have caused
255 trouble for malformed header lines. I hope it is now all correct.
257 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
258 with the "reply" argument non-NULL. The code, however (which originally
259 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
260 but it didn't always do it. This confused somebody who was copying the
261 code for some other use. I have removed all the tests.
263 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
264 feature that was used to support insecure browsers during the U.S. crypto
265 embargo. It requires special client support, and Exim is probably the
266 only MTA that supported it -- and would never use it because real RSA is
267 always available. This code has been removed, because it had the bad
268 effect of slowing Exim down by computing (never used) parameters for the
269 RSA_EXPORT functionality.
271 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
272 authenticator to fail if there's a tab character in the incoming data
273 (there should never be unless someone is messing about, as it's supposed
274 to be base64-encoded). Also added, on Timo's advice, the "secured" option
275 if the connection is using TLS or if the remote IP is the same as the
276 local IP, and the "valid-client-cert option" if a client certificate has
279 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
280 authenticators. This can be used for authorization after authentication
281 succeeds. (In the case of plaintext, it servers for both authentication
284 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
285 if any retry times were supplied.
287 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
288 connection, where there is no client IP address to check. In this
289 situation, the verify now always succeeds.
291 PH/25 Applied John Jetmore's -Mset patch.
293 PH/26 Added -bem to be like -Mset, but loading a message from a file.
295 PH/27 In a string expansion for a processed (not raw) header when multiple
296 headers of the same name were present, leading whitespace was being
297 removed from all of them, but trailing whitespace was being removed only
298 from the last one. Now trailing whitespace is removed from each header
299 before concatenation. Completely empty headers in a concatenation (as
302 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
303 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
305 PH/29 [Removed. This was a change that I later backed out, and forgot to
306 correct the ChangeLog entry (that I had efficiently created) before
307 committing the later change.]
309 PH/30 Exim was sometimes attempting to deliver messages that had suffered
310 address errors (4xx response to RCPT) over the same connection as other
311 messages routed to the same hosts. Such deliveries are always "forced",
312 so retry times are not inspected. This resulted in far too many retries
313 for the affected addresses. The effect occurred only when there were more
314 hosts than the hosts_max_try setting in the smtp transport when it had
315 the 4xx errors. Those hosts that it had tried were not added to the list
316 of hosts for which the message was waiting, so if all were tried, there
317 was no problem. Two fixes have been applied:
319 (i) If there are any address or message errors in an SMTP delivery, none
320 of the hosts (tried or untried) are now added to the list of hosts
321 for which the message is waiting, so the message should not be a
322 candidate for sending over the same connection that was used for a
323 successful delivery of some other message. This seems entirely
324 reasonable: after all the message is NOT "waiting for some host".
325 This is so "obvious" that I'm not sure why it wasn't done
326 previously. Hope I haven't missed anything, but it can't do any
327 harm, as the worst effect is to miss an optimization.
329 (ii) If, despite (i), such a delivery is accidentally attempted, the
330 routing retry time is respected, so at least it doesn't keep
331 hammering the server.
333 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
334 in ${readsocket because some servers need this prod.
336 PH/32 Added some extra debug output when updating a wait-xxx database.
338 PH/33 The hint "could be header name not terminated by colon", which has been
339 given for certain expansion errors for a long time, was not being given
340 for the ${if def:h_colon_omitted{... case.
342 PH/34 The spec says: "With one important exception, whenever a domain list is
343 being scanned, $domain contains the subject domain." There was at least
344 one case where this was not true.
346 PH/35 The error "getsockname() failed: connection reset by peer" was being
347 written to the panic log as well as the main log, but it isn't really
348 panic-worthy as it just means the connection died rather early on. I have
349 removed the panic log writing for the ECONNRESET error when getsockname()
352 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
353 runs only) independently of the message's sender address. This meant
354 that, if the 4xx error was in fact related to the sender, a different
355 message to the same recipient with a different sender could confuse
356 things. In particualar, this can happen when sending to a greylisting
357 server, but other circumstances could also provoke similar problems.
358 I have changed the default so that the retry time for these errors is now
359 based a combination of the sender and recipient addresses. This change
360 can be overridden by setting address_retry_include_sender=false in the
363 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
364 remote server are returned as part of bounce messages. This was not
365 happening for LMTP over a pipe (the lmtp transport), but now it is the
366 same for both kinds of LMTP.
368 PH/38 Despite being documented as not happening, Exim was rewriting addresses
369 in header lines that were in fact CNAMEs. This is no longer the case.
371 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
372 and queue runs started by the daemon processed all messages. This has
373 been fixed so that -R and -S can now usefully be given with -q<time>.
375 PH/40 Import PCRE release 6.7 (fixes some bugs).
377 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
379 PH/42 Give an error if -q is specified more than once.
381 PH/43 Renamed the variables $interface_address and $interface_port as
382 $received_ip_address and $received_port, to make it clear that these
383 values apply to message reception, and not to the outgoing interface when
384 a message is delivered. (The old names remain recognized, of course.)
386 PH/44 There was no timeout on the connect() call when using a Unix domain
387 socket in the ${readsocket expansion. There now is.
389 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
390 be meaningful with "accept".
392 SC/01 Eximstats V1.43
393 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
395 SC/02 Eximstats V1.44
396 Use a glob alias rather than an array ref in the generated
397 parser. This improves both readability and performance.
399 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
400 Collect SpamAssassin and rejection statistics.
401 Don't display local sender or destination tables unless
402 there is data to show.
403 Added average volumes into the top table text output.
405 SC/04 Eximstats V1.46
406 Collect data on the number of addresses (recipients)
407 as well as the number of messages.
409 SC/05 Eximstats V1.47
410 Added 'Message too big' to the list of mail rejection
411 reasons (thanks to Marco Gaiarin).
413 SC/06 Eximstats V1.48
414 Mainlog lines which have GMT offsets and are too short to
415 have a flag are now skipped.
417 SC/07 Eximstats V1.49 (Alain Williams)
418 Added the -emptyok flag.
420 SC/08 Eximstats V1.50
421 Fixes for obtaining the IP address from reject messages.
423 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
424 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
425 whitesspace changes from 4.64-PH/27
427 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
430 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
431 are found, allow negative numbers in numeric criteria)
433 JJ/06 exipick.20061117.2, added new $message_body_missing variable
435 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
436 to match changes made in 4.64-PH/43
438 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
440 PH/47 Put in an explicit test for a DNS lookup of an address record where the
441 "domain" is actually an IP address, and force a failure. This locks out
442 those revolvers/nameservers that support "A-for-A" lookups, in
443 contravention of the specifications.
445 PH/48 When a host name was looked up from an IP address, and the subsequent
446 forward lookup of the name timed out, the host name was left in
447 $sender_host_name, contrary to the specification.
449 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
450 restricted to single-key lookups, Exim was not diagnosing an error if
451 * or *@ was used with a query-style lookup.
453 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
455 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
456 long ago, but noone interested enough thought of it. Let's just say that
457 the "1.1" means that there are some new functions that weren't there at
458 some point in the past.
460 PH/51 Error processing for expansion failure of helo_data from an smtp
461 transport during callout processing was broken.
463 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
464 tested/used via the -bh/-bhc/-bs options.
466 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
467 bug, fixed in subsequent PCRE releases).
469 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
470 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
472 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
478 SC/01 Use a glob alias rather than an array ref in eximstats generated
479 parser. This improves both readability and performance.
481 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
482 Don't display local sender or destination tables in eximstats unless
483 there is data to show.
484 Added average volumes into the eximstats top table text output.
486 SC/03 Collect data on the number of addresses (recipients) as well
487 as the number of messages in eximstats.
489 TF/01 Correct an error in the documentation for the redirect router. Exim
490 does (usually) call initgroups() when daemonizing.
492 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
493 with consistent privilege compared to when running as a daemon.
495 TF/03 Note in the spec that $authenticated_id is not set for local
496 submissions from trusted users.
498 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
499 Thanks to Dean Brooks <dean@iglou.com> for the patch.
501 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
502 by adding some example configuration directives to the default
503 configuration file. A little bit of work is required to uncomment the
504 directives and define how usernames and passwords are checked, but
505 there is now a framework to start from.
507 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
508 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
509 without this. I don't know how relevant this is to other LDAP libraries.
511 PH/02 Add the verb name to the "unknown ACL verb" error.
513 PH/03 Magnus Holmgren's patch for filter_prepend_home.
515 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
517 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
518 directory not expanded when it should be if an expanded home directory
519 was set for the address (which is overridden by the transport).
521 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
524 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
525 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
526 because it is too late at that time, and has no effect.
528 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
529 security issue with \' (bugzilla #107). I could not use the
530 PQescapeStringConn() function, because it needs a PGconn value as one of
533 PH/08 When testing addresses using -bt, indicate those final addresses that
534 are duplicates that would not cause an additional delivery. At least one
535 person was confused, thinking that -bt output corresponded to deliveries.
536 (Suppressing duplicates isn't a good idea as you lose the information
537 about possibly different redirections that led to the duplicates.)
539 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
540 systems where poll() doesn't work, in particular OS X.
542 PH/10 Added more information to debugging output for retry time not reached.
544 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
545 operations in malware.c.
547 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
550 PH/13 If write_rejectlog was set false when logging was sent to syslog with
551 syslog_duplication set false, log lines that would normally be written
552 both the the main log and to the reject log were not written to syslog at
555 PH/14 In the default configuration, change the use of "message" in ACL warn
556 statements to "add_header".
558 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
559 not followed by a command (e.g. "seen endif").
561 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
562 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
565 PH/17 Added extra conditions to the default value of delay_warning_condition
569 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
570 { match{$h_precedence:}{(?i)bulk|list|junk} } \
571 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
574 The Auto-Submitted: and various List- headers are standardised, whereas I
575 don't think Precedence: ever was.
577 PH/18 Refactored debugging code in route_finduser() to show more information,
578 in particular, the error code if getpwnam() issues one.
580 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
581 This is apparently needed in addition to the PH/07 change above to avoid
582 any possible encoding problems.
584 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
585 but not after initializing Perl.
587 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
588 output them only if debugging. By default they are written stderr,
589 apparently, which is not desirable.
591 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
594 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
597 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
599 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
600 authenticated or an ident call has been made. Suppress the default
601 values for $authenticated_id and $authenticated_sender (but permit -oMai
602 and -oMas) when testing with -bh.
604 PH/24 Re-jigged the order of the tests in the default configuration so that the
605 tests for valid domains and recipients precede the DNS black list and CSA
606 tests, on the grounds that those ones are more expensive.
608 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
609 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
610 This bug exists in every version of Exim that I still have, right back to
613 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
614 However, an attempt to turn on case-sensitivity in a regex key by
615 including (?-i) didn't work because the subject string was already
616 lowercased, and the effects were non-intuitive. It turns out that a
617 one-line patch can be used to allow (?-i) to work as expected.
623 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
624 other effects) broke the use of negated acl sub-conditions.
626 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
629 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
630 "Deny" causes Exim to reject the incoming connection with a 554 error.
631 Unfortunately, if there is a major crisis, such as a disk failure,
632 tcp-wrappers gives "deny", whereas what one would like would be some
633 kind of temporary error. A kludge has been added to help with this.
634 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
635 554 error is used if errno is still zero or contains ENOENT (which occurs
636 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
639 PH/03 Add -lutil to the default FreeBSD LIBS setting.
641 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
642 errors. Otherwise a message that provokes a temporary error (when other
643 messages do not) can cause a whole host to time out.
645 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
646 addresses were routed directly to files or pipes from a redirect router.
647 File deliveries just didn't batch; pipe deliveries might have suffered
650 PH/06 A failure to get a lock for a hints database would erroneously always say
651 "Failed to get write lock", even when it was really a read lock.
653 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
654 of 0600. This has been changed to use the value of the lockfile_mode
655 option (which defaults to 0600).
657 PH/08 Applied small patch from the Sieve maintainer.
659 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
660 folder from quota calculations, a direct delivery into this folder messed
661 up the contents of the maildirsize file. This was because the regex was
662 used only to exclude .Trash (or whatever) when the size of the mailbox
663 was calculated. There was no check that a delivery was happening into an
664 excluded directory. This bug has been fixed by ignoring all quota
665 processing for deliveries into excluded directories.
667 PH/10 Added the maildirfolder_create_regex option to appendfile.
673 PH/01 The code for finding all the local interface addresses on a FreeBSD
674 system running IPv6 was broken. This may well have applied to all BSD
675 systems, as well as to others that have similar system calls. The broken
676 code found IPv4 interfaces correctly, but gave incorrect values for the
677 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
678 that it would not match correctly against @[] and not recognize the IPv6
681 PH/02 The ipliteral router was not recognizing addresses of the form user@
682 [ipv6:....] because it didn't know about the "ipv6:" prefix.
684 PH/03 Added disable_ipv6.
686 PH/04 Changed $reply_address to use the raw form of the headers instead of the
687 decoded form, because it is most often used to construct To: headers
688 lines in autoreplies, and the decoded form may well be syntactically
689 invalid. However, $reply_address has leading white space removed, and all
690 newlines turned into spaces so that the autoreply transport does not
693 PH/05 If group was specified without a user on a router, and no group or user
694 was specified on a transport, the group from the router was ignored.
696 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
697 for visible compile-time settings that can be used to change these
698 numbers, for those that want even more. Backwards compatibility with old
699 spool files has been maintained. However, going back to a previous Exim
700 release will lost any variables that are in spool files.
702 PH/07 Two small changes when running in the test harness: increase delay when
703 passing a TCP/IP connection to a new process, in case the original
704 process has to generate a bounce, and remove special handling of
705 127.0.0.2 (sic), which is no longer necessary.
707 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
708 be the same on different OS.
710 PH/09 Moved a debug statement in filter processing to avoid a race problem when
713 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
714 whether --show-vars was specified or not
716 JJ/02 exipick: Added support for new ACL variable spool format introduced
719 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
720 syntactically invalid From: or Reply-to: line, and a filter used this to
721 generate an autoreply, and therefore failed to obtain an address for the
722 autoreply, Exim could try to deliver to a non-existent relative file
723 name, causing unrelated and misleading errors. What now happens is that
724 it logs this as a hard delivery error, but does not attempt to create a
727 PH/11 The exinext utility has a -C option for testing purposes, but although
728 the given file was scanned by exinext itself; it wasn't being passed on
729 when Exim was called.
731 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
732 an end-of-file indication when reading a command response.
734 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
735 compiled. In many other places in Exim, IPv6 addresses are always
736 recognized, so I have changed this. It also means that IPv4 domain
737 literals of the form [IPV4:n.n.n.n] are now always recognized.
739 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
740 used if the router is not running as root, for example, when verifying at
741 ACL time, or when using -bh. The debugging output from this situation was
742 non-existent - all you got was a failure to exec. I have made two
745 (a) Failures to set uid/gid, the current directory, or a process leader
746 in a subprocess such as that created by queryprogram now generate
747 suitable debugging ouput when -d is set.
749 (b) The queryprogram router detects when it is not running as root,
750 outputs suitable debugging information if -d is set, and then runs
751 the subprocess without attempting to change uid/gid.
753 PH/15 Minor change to Makefile for building test_host (undocumented testing
756 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
757 additional section of a DNS packet that returns MX or SRV records.
758 Instead, it always explicitly searches for A/AAAA records. This avoids
759 major problems that occur when a DNS server includes only records of one
760 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
761 fixed another bug: if SRV records were looked up and the corresponding
762 address records were *not* found in the additional section, the port
763 values from the SRV records were lost.
765 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
766 using the correct key (the original address) when searching the retry
767 rules in order to find which one to use for generating the retry hint.
769 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
770 adding the default one. Similarly, if it contains a Reply-To: header, the
771 errors_reply_to option, if set, is not used.
773 PH/19 When calculating a retry time, Exim used to measure the "time since
774 failure" by looking at the "first failed" field in the retry record. Now
775 it does not use this if it is later than than the arrival time of the
776 message. Instead it uses the arrival time. This makes for better
777 behaviour in cases where some deliveries succeed, thus re-setting the
778 "first failed" field. An example is a quota failure for a huge message
779 when small messages continue to be delivered. Without this change, the
780 "time since failure" will always be short, possible causing more frequent
781 delivery attempts for the huge message than are intended.
782 [Note: This change was subsequently modified - see PH/04 for 4.62.]
784 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
785 $1, $2, $3) because the numerical variables can be reset during some
786 expansion items (e.g. "match"), thereby losing the authentication data.
788 PH/21 Make -bV show the size of off_t variables so that the test suite can
789 decide whether to run tests for quotas > 2G.
791 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
792 mailbox_size, and mailbox_filecount in the appendfile transport. If a
793 filecount value is greater than 2G or if a quota value is greater than 2G
794 on a system where the size of off_t is not greater than 4, a panic error
797 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
798 never match. The debug and -bh output now contains an explicit error
799 message indicating a malformed IPv4 address or mask.
801 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
802 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
805 PH/25 Do not write to syslog when running in the test harness. The only
806 occasion when this arises is a failure to open the main or panic logs
807 (for which there is an explicit test).
809 PH/26 Added the /no_tell option to "control=freeze".
811 PH/27 If a host name lookup failed very early in a connection, for example, if
812 the IP address matched host_lookup and the reverse lookup yielded a name
813 that did not have a forward lookup, an error message of the form "no IP
814 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
815 could be logged. Now it outputs the IP address instead of "NULL".
817 PH/28 An enabling patch from MH: add new function child_open_exim2() which
818 allows the sender and the authenticated sender to be set when
819 submitting a message from within Exim. Since child_open_exim() is
820 documented for local_scan(), the new function should be too.
822 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
823 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
824 results in an empty string is now treated as unset.
826 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
828 PH/31 Added sender_verify_fail logging option.
830 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
831 needed by RFC 822 but not by RFC 2822 was commented out. I have now
832 tidied the source and removed it altogether.
834 PH/33 When a queue run was abandoned because the load average was too high, a
835 log line was always written; now it is written only if the queue_run log
836 selector is set. In addition, the log line for abandonment now contains
837 information about the queue run such as the pid. This is always present
838 in "start" and "stop" lines but was omitted from the "abandon" line.
840 PH/34 Omit spaces between a header name and the colon in the error message that
841 is given when verify = headers_syntax fails (if there are lots of them,
842 the message gets confusing).
844 PH/35 Change the default for dns_check_names_pattern to allow slashes within
845 names, as there are now some PTR records that contain slashes. This check
846 is only to protect against broken name servers that fall over on strange
847 characters, so the fact that it applies to all lookups doesn't matter.
849 PH/36 Now that the new test suite is complete, we can remove some of the
850 special code in Exim that was needed for the old test suite. For example,
851 sorting DNS records because real resolvers return them in an arbitrary
852 order. The new test suite's fake resolver always returns records in the
855 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
856 bounces) except when queue_only is set, to avoid logging races between
857 the different processes.
859 PH/38 Panic-die if .include specifies a non-absolute path.
861 PH/39 A tweak to the "H" retry rule from its user.
863 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
864 a label. They prevented compilation on older perls.
866 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
867 a warning to be raised on newish perls.
869 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
870 on queue. Changes to match documented behaviour of showing count of
871 messages matching specified criteria.
873 PH/40 Changed the default ident timeout from 30s to 5s.
875 PH/41 Added support for the use of login_cap features, on those BSD systems
876 that have them, for controlling the resources used by pipe deliveries.
878 PH/42 The content-scanning code uses fopen() to create files in which to put
879 message data. Previously it was not paying any attention to the mode of
880 the files. Exim runs with umask(0) because the rest of the code creates
881 files with open(), and sets the required mode explicitly. Thus, these
882 files were ending up world-writeable. This was not a big issue, because,
883 being within the spool directory, they were not world-accessible. I have
884 created a function called modefopen, which takes an additional mode
885 argument. It sets umask(777), creates the file, chmods it to the required
886 mode, then resets the umask. All the relevant calls to fopen() in the
887 content scanning code have been changed to use this function.
889 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
890 to 24 hours. This avoids potential overflow problems when processing G
891 and H retry rules. I suspect nobody ever tinkers with this value.
893 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
895 PH/45 When the plaintext authenticator is running as a client, the server's
896 challenges are checked to ensure they are valid base64 strings. By
897 default, the authentication attempt is cancelled if an invalid string is
898 received. Setting client_ignore_invalid_base64 true ignores these errors.
899 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
900 they are received. Thus, the responses can be made to depend on the
901 challenges. If an invalid string is ignored, an empty string is placed in
904 PH/46 Messages that are created by the autoreply transport now contains a
905 References: header, in accordance with RFCs 2822 and 3834.
907 PH/47 Added authenticated_sender_force to the smtp transport.
909 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
911 PH/49 Installed latest patch from the Sieve maintainer.
913 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
914 was also set, the appendfile transport was unnecessarily scanning a
915 directory of message files (e.g. for maildir delivery) to find the count
916 of files (along with the size), even though it did not need this
917 information. It now does the scan only if it needs to find either the
918 size of the count of files.
920 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
922 PH/52 Two bugs concerned with error handling when the smtp transport is
925 (i) Exim was not creating retry information for temporary errors given
926 for individual recipients after the DATA command when the smtp transport
927 was used in LMTP mode. This meant that they could be retried too
928 frequently, and not timed out correctly.
930 (ii) Exim was setting the flag that allows error details to be returned
931 for LMTP errors on RCPT commands, but not for LMTP errors for individual
932 recipients that were returned after the DATA command.
934 PH/53 This is related to PH/52, but is more general: for any failing address,
935 when detailed error information was permitted to be returned to the
936 sender, but the error was temporary, then after the final timeout, only
937 "retry timeout exceeded" was returned. Now it returns the full error as
938 well as "retry timeout exceeded".
940 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
941 do this, and (what is worse) MTAs that accept it.
943 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
944 will now be deprecated.
946 PH/56 New os.c-cygwin from the Cygwin maintainer.
948 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
949 formats (previously only available in exim formats via -bpr, -bpru,
950 and -bpra. Now also available in native and exiqgrep formats)
952 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
953 with very large, slow to parse queues
955 JJ/08 exipick: added ! as generic prefix to negate any criteria format
957 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
959 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
960 responses to authentication challenges, though it was showing the
961 challenges; (ii) I've removed the CR characters from the debug output for
964 PH/58 Allow for the insertion of a newline as well as a space when a string
965 is turned into more than one encoded-word during RFC 2047 encoding. The
966 Sieve code now uses this.
968 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
969 data_4xx, lost_connection, tls_required.
971 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
972 message was being sent as an SMTP response.
974 PH/61 Add -l and -k options to exicyclog.
976 PH/62 When verifying, if an address was redirected to one new address, so that
977 verification continued, and the new address failed or deferred after
978 having set something in $address_data, the value of $address_data was not
979 passed back to the ACL. This was different to the case when no
980 redirection occurred. The value is now passed back in both cases.
982 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
983 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
984 use login_cap.h, so on its own it isn't the distinguishing feature. The
985 new name refers directly to the setclassresources() function.
987 PH/65 Added configuration files for NetBSD3.
989 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
991 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
992 is preferred over IPv4.
994 PH/68 The bounce_return_message and bounce_return_body options were not being
995 honoured for bounces generated during the reception of non-SMTP messages.
996 In particular, this applied to messages rejected by the ACL. This bug has
997 been fixed. However, if bounce_return_message is true and bounce_return_
998 body is false, the headers that are returned for a non-SMTP message
999 include only those that have been read before the error was detected.
1000 (In the case of an ACL rejection, they have all been read.)
1002 PH/69 The HTML version of the specification is now built in a directory called
1003 spec_html instead of spec.html, because the latter looks like a path with
1004 a MIME-type, and this confuses some software.
1006 PH/70 Catch two compiler warnings in sieve.c.
1008 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1009 function verify_get_ident() calls ip_connect() to connect a socket, but
1010 if the "connect()" function timed out, ip_connect() used to close the
1011 socket. However, verify_get_ident() also closes the socket later, and in
1012 between Exim writes to the log, which may get opened at this point. When
1013 the socket was closed in ip_connect(), the log could get the same file
1014 descriptor number as the socket. This naturally causes chaos. The fix is
1015 not to close the socket in ip_connect(); the socket should be closed by
1016 the function that creates it. There was only one place in the code where
1017 this was missing, in the iplookup router, which I don't think anybody now
1018 uses, but I've fixed it anyway.
1020 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1021 well as to direct DNS lookups. Otherwise the handling of names in host
1022 lists is inconsistent and therefore confusing.
1028 PH/01 Two changes to the default runtime configuration:
1030 (1) Move the checks for relay_from_hosts and authenticated clients from
1031 after to before the (commented out) DNS black list checks.
1033 (2) Add control=submission to the relay_from_hosts and authenticated
1034 clients checks, on the grounds that messages accepted by these
1035 statements are most likely to be submissions.
1037 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1039 (1) Generate an error if the third argument for the ${prvs expansion is
1042 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1045 (3) Reset the variables that are obtained from the first argument of
1046 ${prvscheck and used in the second argument before leaving the code,
1047 because their memory is reclaimed, so using them afterwards may do
1050 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1051 one (it's much easier than Tom thought :-).
1053 (5) Because of (4), we can now allow for the use of $prvscheck_result
1054 inside the third argument.
1056 PH/03 For some reason, the default setting of PATH when running a command from
1057 a pipe transport was just "/usr/bin". I have changed it to
1060 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1061 anything to be listed in the output from -bV.
1063 PH/05 When a filter generated an autoreply, the entire To: header line was
1064 quoted in the delivery log line, like this:
1066 => >A.N.Other <ano@some.domain> <original@ddress> ...
1068 This has been changed so that it extracts the operative address. There
1069 may be more than one such address. If so, they are comma-separated, like
1072 => >ano@some.domain,ona@other.domain <original@ddress> ...
1074 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1075 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1076 not being looked up in the rDNS to get a host name, Exim was showing the
1077 IP address twice in Received: lines, even though the IP addresses were
1078 identical. For example:
1080 Received: from [1.2.3.4] (helo=[1.2.3.4])
1082 However, if the real host name was known, it was omitting the HELO data
1083 if it matched the actual IP address. This has been tidied up so that it
1084 doesn't show the same IP address twice.
1086 PH/07 When both +timestamp and +memory debugging was on, the value given by
1087 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1088 called by the debug printing, thereby overwriting the timestamp buffer.
1089 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1091 PH/08 When the original message was included in an autoreply transport, it
1092 always said "this is a copy of the message, including all the headers",
1093 even if body_only or headers_only was set. It now gives an appropriate
1096 PH/09 Applied a patch from the Sieve maintainer which:
1098 o fixes some comments
1099 o adds the (disabled) notify extension core
1100 o adds some debug output for the result of if/elsif tests
1101 o points to the current vacation draft in the documentation
1102 and documents the missing references header update
1106 o fixes a bug in processing the envelope test (when testing
1107 multiple envelope elements, the last element determinted the
1110 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1111 Electronic Mail") by including:
1113 Auto-submitted: auto-generated
1115 in the messages that it generates (bounce messages and others, such as
1116 warnings). In the case of bounce messages for non-SMTP mesages, there was
1117 also a typo: it was using "Auto_submitted" (underscore instead of
1118 hyphen). Since every message generated by Exim is necessarily in response
1119 to another message, thes have all been changed to:
1121 Auto-Submitted: auto-replied
1123 in accordance with these statements in the RFC:
1125 The auto-replied keyword:
1127 - SHOULD be used on messages sent in direct response to another
1128 message by an automatic process,
1130 - MUST NOT be used on manually-generated messages,
1132 - MAY be used on Delivery Status Notifications (DSNs) and Message
1133 Disposition Notifications (MDNs),
1135 - MUST NOT be used on messages generated by automatic or periodic
1136 processes, except for messages which are automatic responses to
1139 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1140 to the default Received: header definition.
1142 PH/12 Added log selector acl_warn_skipped (default on).
1144 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1145 variables because (a) they are in the wrong storage pool and (b) even if
1146 they were copied, it wouldn't work properly because of the caching.
1148 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1149 checking when decoding. Apparently there are clients that generate
1150 overlong encoded strings. Why am I not surprised?
1152 PH/15 If the first argument of "${if match_address" was not empty, but did not
1153 contain an "@" character, Exim crashed. Now it writes a panic log message
1154 and treats the condition as false.
1156 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1158 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1159 extension "envelope-auth". The code is finished and in agreement with
1160 other implementations, but there is no documentation so far and in fact,
1161 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1162 not changing the active code.
1164 Print executed "if" and "elsif" statements when debugging is used. This
1165 helps a great deal to understand what a filter does.
1167 Document more things not specified clearly in RFC3028. I had all this
1168 sorted out, when out of a sudden new issues came to my mind. Oops."
1170 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1173 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1174 canonical form (as documented). However, after a host name lookup from
1175 the IP address, check_host() was doing a simple string comparison with
1176 addresses acquired from the DNS when checking that the found name did
1177 have the original IP as one of its addresses. Since any found IPv6
1178 addresses are likely to be in abbreviated form, the comparison could
1179 fail. Luckily, there already exists a function for doing the comparison
1180 by converting both addresses to binary, so now that is used instead of
1181 the text comparison.
1183 PH/20 There was another similar case to PH/19, when a complete host name was
1184 given in a host list; looking up its IP address could give an abbreviated
1185 form, whereas the current host's name might or might not be abbreviated.
1186 The same fix has been applied.
1192 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1193 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1196 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1197 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1199 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1201 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1202 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1203 Also, for "Auto-Submitted", treat anything other than "no" as
1204 non-personal, in accordance with RFC 3834. (Previously it treated
1205 anything starting "auto-" as non-personal.)
1207 TF/01 The control=submission/name=... option had a problem with syntax
1208 errors if the name included a slash character. The /name= option
1209 now slurps the rest of the string, so it can include any characters
1210 but it must come last in the list of options (after /sender_retain
1213 PH/05 Some modifications to the interface to the fake nameserver for the new
1221 TK/01 Added the "success_on_redirect" address verification option. See
1222 NewStuff for rationale and an example.
1224 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1226 PH/02 Patch to exigrep to allow it to work on syslog lines.
1228 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1229 fread() to skip over the body file's header line, because in Cygwin the
1230 header line is locked and is inaccessible.
1232 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1233 co-exist for some time) to make it clear that it is the Exim ID that is
1234 referenced, not the Message-ID: header line.
1236 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1237 string_format() function, because snprintf() does not exist on all
1240 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1241 ${stat: expansion item.
1243 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1244 protocol synchronization error", to keep the pedants happy.
1246 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1247 well as for IRIX systems, when gcc is being used. See the host.c source
1250 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1252 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1255 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1256 transport and to the smtp transport in LMTP mode.
1258 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1260 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1261 run a filter in a subprocess. This could lead to confusion in subsequent
1262 lookups in the parent process. There should also be a search_tidyup() at
1263 the end of the subprocess.
1265 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1266 only if the host matched helo_try_verify_hosts, which caused the
1267 verification to occur when the EHLO/HELO command was issued. The ACL just
1268 tested the remembered result. Now, if a previous verification attempt has
1269 not happened, "verify = helo" does it there and then.
1271 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1273 TK/03 Fix log output including CR from clamd.
1275 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1276 exist provoked a memory error which could cause a segfault.
1278 PH/15 Installed PCRE 6.2
1280 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1282 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1283 of the problem. Specifically, suggested +O2 rather than +O1 for the
1286 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1288 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1289 record, the port was not correctly logged when the outgoing_port log
1290 selector was set (it logged the transort's default port).
1292 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1293 fallback_hosts, and "hosts" in the smtp transport.
1295 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1296 host errors such as "Connection refused".
1298 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1299 authentication with radiusclient 0.4.9:
1301 - Error returned from rc_read_config was caught wrongly
1302 - Username/password not passed on to radius server due to wrong length.
1304 The presumption is that some radiusclient API changes for 4.51/PH/17
1305 were not taken care of correctly. The code is still untested by me (my
1306 Linux distribution still has 0.3.2 of radiusclient), but it was
1307 contributed by a Radius user.
1309 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1310 expanding the "port" option of the smtp transport.
1312 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1313 while reading a MIME header. Thanks to Tom Hughes for a patch.
1315 PH/24 Include config.h inside local_scan.h so that configuration settings are
1318 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1319 that in an ACL for RCPT (for example), you can examine exactly what was
1322 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1323 commands, but it was not correctly comparing the address with the actual
1324 client host address. Thus, it would show the EHLO address in Received:
1325 header lines when this was not necessary.
1327 PH/27 Added the % operator to ${eval:}.
1329 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1330 it should be ignoring failures (because with -C, for example, it has lost
1331 privilege). It wasn't ignoring creation failures other than "already
1334 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1337 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1338 because an input line was too long, either on its own, or by virtue of
1339 too many continuations, the temporary file was not being removed, and the
1340 return code was incorrect.
1342 PH/31 Missing "BOOL" in function definition in filtertest.c.
1344 PH/32 Applied Sieve patches from the maintainer.
1346 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1348 PH/33 Added "verify = not_blind".
1350 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1351 Local/Makefile (with some defaults set). These are used in built scripts
1352 such as exicyclog, but they have never been used in the exim_install
1353 script (though there are many overriding facilities there). I have
1354 arranged that the exim_install script now takes note of these two
1357 PH/35 Installed configuration files for Dragonfly.
1359 PH/36 When a locally submitted message by a trusted user did not contain a
1360 From: header, and the sender address was obtained from -f or from an SMTP
1361 MAIL command, and the trusted user did not use -F to supply a sender
1362 name, $originator_name was incorrectly used when constructing a From:
1363 header. Furthermore, $originator_name was used for submission mode
1364 messages from external hosts without From: headers in a similar way,
1365 which is clearly wrong.
1367 PH/37 Added control=suppress_local_fixups.
1369 PH/38 When log_selector = +received_sender was set, and the addition of the
1370 sender made the log line's construction buffer exactly full, or one byte
1371 less than full, an overflow happened when the terminating "\n" was
1374 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1375 when the result of a list match is failure because a DNS lookup failed.
1377 PH/40 RM_COMMAND is now used in the building process.
1379 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1380 the "build-* directories that it finds.
1382 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1383 address in a domain literal was a prefix of an interface address.
1385 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1386 when verifying a sender address, unless rewrite_headers is false.
1388 PH/44 Wrote a long comment about why errors_to addresses are verified as
1389 recipients, not senders.
1391 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1392 the ratelimit ACL was added.
1394 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1396 PH/46 Added extra information about PostgreSQL errors to the error string.
1398 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1399 avoiding the need to install special zones in a real server. This is
1400 backwards compatible; if it can't find the fake resolver, it drops back.
1401 Thus, both old and new test suites can be run.
1403 TF/02 Added util/ratelimit.pl
1405 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1406 clock is set back in time.
1408 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1409 Candler <B.Candler@pobox.com>.
1411 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1412 OK for addresses that are the result of redirections.
1414 PH/48 A number of further additions for the benefit of the new test suite,
1415 including a fake gethostbyname() that interfaces to the fake DNS resolver
1418 TF/06 The fix for widen_domains has also been applied to qualify_single and
1419 search_parents which are the other dnslookup options that can cause
1422 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1425 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1427 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1428 service" if it encountered an MX record with an empty target hostname.
1429 The message is now "an MX or SRV record indicated no SMTP service".
1431 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1432 if the DNS of the sending site is misconfigured. This is quite a
1433 common situation. This change restores the behaviour of treating a
1434 helo verification defer as a failure.
1436 PH/51 If self=fail was set on a router, the bounce message did not include the
1437 actual error message.
1443 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1445 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1446 command itself ended in error, the underlying message about the transport
1447 filter timeout was being overwritten with the pipe command error. Now the
1448 underlying error message should be appended to the second error message.
1450 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1452 PH/02 Reduce the amount of output that "make" produces by default. Full output
1453 can still be requested.
1455 PH/03 The warning log line about a condition test deferring for a "warn" verb
1456 was being output only once per connection, rather than after each
1457 occurrence (because it was using the same function as for successful
1458 "warn" verbs). This seems wrong, so I have changed it.
1460 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1461 it should not have, which might have caused a crash in the right
1462 circumstances, but probably never did.
1464 PH/04 Installed a modified version of Tony Finch's patch to make submission
1465 mode fix the return path as well as the Sender: header line, and to
1466 add a /name= option so that you can make the user's friendly name appear
1469 TF/03 Added the control = fakedefer ACL modifier.
1471 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1472 Mark Lowes for thorough testing.
1474 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1476 TK/03 Merged latest SRS patch from Miles Wilton.
1478 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1479 with the definition in sysexits.h (which is #included earlier).
1480 Fortunately, Exim does not actually use EX_OK. The code used to try to
1481 preserve the sysexits.h value, by assumimg that macro definitions were
1482 scanned for macro replacements. I have been disabused of this notion,
1483 so now the code just undefines EX_OK before #including unistd.h.
1485 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1486 in the smtp transport. When a block could not be written in a single
1487 write() function, the timeout was being re-applied to each part-write.
1488 This seems wrong - if the receiver was accepting one byte at a time it
1489 would take for ever. The timeout is now adjusted when this happens. It
1490 doesn't have to be particularly precise.
1492 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1493 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1495 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1496 without a domain if the check to <postmaster@domain> fails.
1498 SC/01 Eximstats: added -xls and the ability to specify output files
1499 (patch written by Frank Heydlauf).
1501 SC/02 Eximstats: use FileHandles for outputing results.
1503 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1505 SC/04 Eximstats: fixed display of large numbers with -nvr option
1507 SC/05 Eximstats: fixed merging of reports with empty tables.
1509 SC/06 Eximstats: added the -include_original_destination flag
1511 SC/07 Eximstats: removed tabs and trailing whitespace.
1513 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1515 TK/06 MBOX spool code: Add real "From " MBOX separator line
1516 so the .eml file is really in mbox format (even though
1517 most programs do not really care). Patch from Alex Miller.
1519 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1520 The latter is generated from $received_to and is only set if the
1521 message has one envelope recipient. SA can use these headers,
1522 obviously out-of-the-box. Patch from Alex Miller.
1524 PH/08 The ${def test on a variable was returning false if the variable's
1525 value was "0", contrary to what the specification has always said!
1526 The result should be true unless the variable is empty.
1528 PH/09 The syntax error of a character other than { following "${if
1529 def:variable_name" (after optional whitespace) was not being diagnosed.
1530 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1531 accidental colon was present, for example, could give incorrect results.
1533 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1534 result is used (not including appendfile, where other changes are about
1537 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1538 This involved changing a lot of size variables from int to off_t. It
1539 should work with maildirs and everything.
1541 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1542 spamd dying while we are connected to it.
1544 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1547 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1548 function for BDB 4.3.
1550 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1552 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1553 the sources for PCRE have been re-arranged and more files are now
1556 PH/15 The code I had for printing potentially long long variables in PH/11
1557 above was not the best (it lost precision). The length of off_t variables
1558 is now inspected at build time, and an appropriate printing format (%ld
1559 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1560 to be "long long int" or "long int". This is needed for the internal
1561 formatting function string_vformat().
1563 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1564 the configuration file to be ":syslog", then the script "guesses" where
1565 the logs files are, rather than using the compiled in default. In our
1566 case the guess is not the same as the compiled default, so the script
1567 suddenly stopped working when I started to use syslog. The patch checks
1568 to see if log_file_path is "". If so, it attempts to read it from exim
1569 with no configuration file to get the compiled in version, before it
1570 falls back to the previous guessing code."
1572 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1573 implementing BATV in an Exim configuration. See NewStuff for the gory
1576 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1577 Makefile that are specific to HP-UX.
1579 PH/18 If the "use_postmaster" option was set for a recipient callout together
1580 with the "random" option, the postmaster address was used as the MAIL
1581 FROM address for the random test, but not for the subsequent recipient
1582 test. It is now used for both.
1584 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1585 patch removes a few documentation additions to RFC 3028, because the
1586 latest draft now contains them. It adds the new en;ascii-case comparator
1587 and a new error check for 8bit text in MIME parts. Comparator and
1588 require names are now matched exactly. I enabled the subaddress
1589 extension, but it is not well tested yet (read: it works for me)."
1591 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1592 rework some of the code of TK/09 above to avoid the hardwired use of
1593 "%lld" and "long long". Replaced the call to snprintf() with a call to
1596 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1597 records point to non-existent hosts", "retry timeout exceeded", and
1598 "retry time not reached for any host after a long failure period".
1600 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1601 experimental DomainKeys support:
1603 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1604 (2) On an error such as an illegally used "control", the wrong name for
1605 the control was given.
1607 These problems did NOT occur unless DomainKeys support was compiled.
1609 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1611 PH/24 Added ${if match_ip condition.
1613 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1614 databases so that it will be absolutely obvious if a crash occurs in the
1615 DB library. This is a regular occurrence (often caused by mis-matched
1618 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1619 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1620 picked up on a user's system that detects such things. There doesn't seem
1621 to be a gcc warning option for this - only an attribute that has to be
1622 put on the function's prototype. It seems that in Fedora Core 4 they have
1623 set this on a number of new functions. No doubt there will be more in due
1626 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1627 specify a transport. However, if an address that was verified by such a
1628 router was the subject of a callout, Exim crashed because it tried to
1629 read the rcpt_include_affixes from the non-existent transport. Now it
1630 just assumes that the setting of that option is false. This bug was
1631 introduced by 4.51/PH/31.
1633 PH/28 Changed -d+all to exclude +memory, because that information is very
1634 rarely of interest, but it makes the output a lot bigger. People tend to
1635 do -d+all out of habit.
1637 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1638 code in os-type was giving problems when libc.so lives in lib64, like on
1641 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1642 aren't the modern standard, and it seems that some systems' include files
1643 don't always have them. Exim was already checking for some of the newer
1644 ones like T_AAAA, and defining it itself. I've added checks for all the
1645 record types that Exim uses.
1647 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1648 not automatically generating a new one, as it is supposed to. This
1649 prevented TLS from working. If the file did exist, but contained invalid
1650 data, a new version was generated, as expected. It was only the case of a
1651 non-existent file that was broken.
1653 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1654 with a change in libdomainkeys > 0.64.
1656 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1657 from DNS. If the selector record carries the flag, it now has
1658 precedence over the domain-wide flag.
1660 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1662 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1663 the use of a port name that isn't defined in /etc/services) occurred, the
1664 message was deferred as in a normal delivery, and thus remained on the
1665 spool, instead of being failed because of the mua_wrapper setting. This
1666 is now fixed, and I tidied up some of the mua_wrapper messages at the
1669 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1670 the messages in a hash of arrays rather than using individual hashes.
1671 This is a bit cleaner and results in dramatic memory savings, albeit
1672 at a slight CPU cost.
1674 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1675 as requested by Marc Sherman.
1677 SC/10 Eximstats: added histograms for user specified patterns as requested
1680 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1682 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1683 fopen() in the content-scanning modules that did not already have it.
1689 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1690 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1692 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1694 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1696 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1697 to test Sieve filters that use "vacation".
1699 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1700 that changes the way the GnuTLS parameters are stored in the cache file.
1701 The new format can be generated externally. For backward compatibility,
1702 if the data in the cache doesn't make sense, Exim assumes it has read an
1703 old-format file, and it generates new data and writes a new file. This
1704 means that you can't go back to an older release without removing the
1707 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1708 work if there are any delivery delays because "one_time" forces the
1709 parent to be marked "delivered", so its unseen clone is never tried
1710 again. For this reason, Exim now forbids the simultaneous setting of
1713 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1714 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1715 entry if you want to know the details. The fix, however, neglected to
1716 consider the case when local delivery batching is involved. The test for
1717 "previously delivered" was not happening when checking to see if an
1718 address could be batched with a previous (undelivered) one; under
1719 certain circumstances this could lead to multiple deliveries to the same
1722 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1723 in its include files, and this causes problems building Exim.
1725 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1726 header_syntax) but Exim was just ignoring anything given after a slash.
1727 In particular, this caused confusion with an attempt to use "verify =
1728 reverse_host_lookup/defer_ok". An error is now given when options are
1729 supplied for verify items that do not have them. (Maybe reverse_host_
1730 lookup should have a defer_ok option, but that's a different point.)
1732 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1733 defined by RFC 821) to 2048, because there were problems with some AUTH
1734 commands, and RFC 1869 says the size should be increased for extended
1735 SMTP commands that take arguments.
1737 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1740 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1741 "unknown" error; now it says that the functionality isn't in the binary.
1743 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1744 an address' error message when a string expansion fails (syntax or
1745 whatever). Otherwise the password may appear in the log. Following change
1746 PH/42 below, there is no longer a chance of it appearing in a bounce
1749 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1751 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1752 of its addresses was ever tried. (Bugzilla bug #2.)
1754 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1755 the result incorrectly in the debug output. (It correctly added a newline
1756 to what was transported.)
1758 TF/01 Added $received_time.
1760 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1761 commented out examples of how to interface to a virus scanner and to
1762 SpamAssassin. Also added commented examples of av_scanner and
1763 spamd_address settings.
1765 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1766 and controls are allowed in which ACLs. There were a couple of minor
1767 errors. Some of the entries in the conditions table (which is a table of
1768 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1769 negation of where the condition IS allowed.
1771 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1773 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1774 header file does not have a version number, so I've had to invent a new
1775 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1776 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1777 radiusclient), but it was contributed by a Radius user.
1779 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1780 files or directories, for OpenSSL.
1782 PH/19 When an Exim process that is running as root has to create an Exim log
1783 file, it does so in a subprocess that runs as exim:exim so as to get the
1784 ownership right at creation (otherwise, other Exim processes might see
1785 the file with the wrong ownership). There was no test for failure of this
1786 fork() call, which would lead to the process getting stuck as it waited
1787 for a non-existent subprocess. Forks do occasionally fail when resources
1788 run out. I reviewed all the other calls to fork(); they all seem to check
1791 PH/20 When checking for unexpected SMTP input at connect time (before writing
1792 the banner), Exim was not dealing correctly with a non-positive return
1793 from the read() function. If the client had disconnected by this time,
1794 the result was a log entry for a synchronization error with an empty
1795 string after "input=" when read() returned zero. If read() returned -1
1796 (an event I could not check), uninitialized data bytes were printed.
1797 There were reports of junk text (parts of files, etc) appearing after
1800 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1802 PH/22 Added support for macro redefinition, and (re)definition in between
1803 driver and ACL definitions.
1805 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1806 forgetting to use the resulting value; it was using the unexpanded value.
1808 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1809 hadn't been configured. The fix is from Juergen Kreileder, who
1810 understands it better than I do:
1812 "Here's what I see happening with three configured cyrus_sasl
1813 authenticators configured (plain, login, cram-md5):
1815 On startup auth_cyrus_sasl_init() gets called for each of these.
1816 This means three calls to sasl_listmech() without a specified mech_list.
1817 => SASL tests which mechs of all available mechs actually work
1818 => three warnings about OTP not working
1819 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1821 With the patch, sasl_listmech() also gets called three times. But now
1822 SASL's mech_list option is set to the server_mech specified in the the
1823 authenticator. Or in other words, the answer from sasl_listmech()
1824 gets limited to just the mech you're testing for (which is different
1826 => the return list contains just 'plain' or 'login', 'cram-md5' or
1827 nothing depending on the value of ob->server_mech.
1829 I've just tested the patch: Authentication still works fine,
1830 unavailable mechs specified in the exim configuration are still
1831 caught, and the auth.log warnings about OTP are gone."
1833 PH/25 When debugging is enabled, the contents of the command line are added
1834 to the debugging output, even when log_selector=+arguments is not
1837 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1838 answer is "GNU", and only if the return is "GNU/something" is the answer
1841 PH/27 $acl_verify_message is now set immediately after the failure of a
1842 verification in an ACL, and so is available in subsequent modifiers. In
1843 particular, the message can be preserved by coding like this:
1845 warn !verify = sender
1846 set acl_m0 = $acl_verify_message
1848 Previously, $acl_verify_message was set only while expanding "message"
1849 and "log_message" when a very denied access.
1851 PH/28 Modified OS/os.c-Linux with
1853 -#ifndef OS_LOAD_AVERAGE
1854 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1856 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1857 nomenclature these days.)
1859 PH/29 Installed patch from the Sieve maintainer that adds the options
1860 sieve_useraddress and sieve_subaddress to the redirect router.
1862 PH/30 In these circumstances:
1863 . Two addresses routed to the same list of hosts;
1864 . First host does not offer TLS;
1865 . First host accepts first address;
1866 . First host gives temporary error to second address;
1867 . Second host offers TLS and a TLS session is established;
1868 . Second host accepts second address.
1869 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1870 and peerdn, if requested) that were in fact used only for the second
1873 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1874 attention to any local part prefix or suffix that was matched by the
1875 router that accepted the address. It now behaves in the same way as it
1876 does for delivery: the affixes are removed from the local part unless
1877 rcpt_include_affixes is set on the transport.
1879 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1880 timeout during the DATA phase of an incoming message.
1882 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1883 applied a patch sanctioned by the Sieve maintainer.
1885 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1886 the uid or gid is negative. A case of a negative gid caused this to be
1887 noticed. The fix allows for either to be negative.
1889 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1890 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1893 PH/36 The change PH/12 above was broken. Fixed it.
1895 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1896 the grounds that routing the same address twice would always produce the
1897 same answer. This might have been true once, but it is certainly no
1898 longer true now. Routing a child address may depend on the previous
1899 routing that produced that child. Some complicated redirection strategies
1900 went wrong when messages had multiple recipients, and made Exim's
1901 behaviour dependent on the order in which the addresses were given.
1903 I have moved the duplicate checking until after the routing is complete.
1904 Exim scans the addresses that are assigned to local and remote
1905 transports, and removes any duplicates. This means that more work will be
1906 done, as duplicates will always all be routed, but duplicates are
1907 presumably rare, so I don't expect this is of any significance.
1909 For deliveries to pipes, files, and autoreplies, the duplicate checking
1910 still happens during the routing process, since they are not going to be
1913 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1914 It corrects a timeout issue with spamd. This is Ian's comment: "The
1915 background is that sometimes spamd either never reads data from a
1916 connection it has accepted, or it never writes response data. The exiscan
1917 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1918 blindly assumes that writes won't block so it may never time out."
1920 PH/39 Allow G after quota size as well as K and M.
1922 PH/40 The value set for $authenticated_id in an authenticator may not contain
1923 binary zeroes or newlines because the value is written to log lines and
1924 to spool files. There was no check on this. Now the value is run through
1925 the string_printing() function so that such characters are converted to
1926 printable escape sequences.
1928 PH/41 $message_linecount is a new variable that contains the total number of
1929 lines in the message. Compare $body_linecount, which is the count for the
1932 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1933 bounce and delay warning messages, except in certain special cases, which
1936 (a) An SMTP error message from a remote host;
1937 (b) A message specified in a :fail: redirection;
1938 (c) A message specified in a "fail" command in a system filter;
1939 (d) A message specified in a FAIL return from the queryprogram router;
1940 (e) A message specified by the cannot_route_message router option.
1942 In these cases only, Exim does include the error details in bounce and
1943 warning messages. There are also a few cases where bland messages such
1944 as "unrouteable address" or "local delivery error" are given.
1946 PH/43 $value is now also set for the "else" part of a ${run expansion.
1948 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1949 being worked on, but at least Exim now implements the latest version to
1952 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1953 process to complete was treated as a delivery failure, a timeout while
1954 writing the message to the pipe was logged, but erroneously treated as a
1955 successful delivery. Such timeouts include transport filter timeouts. For
1956 consistency with the overall process timeout, these timeouts are now
1957 treated as errors, giving rise to delivery failures by default. However,
1958 there is now a new Boolean option for the pipe transport called
1959 timeout_defer, which, if set TRUE, converts the failures into defers for
1960 both kinds of timeout. A transport filter timeout is now identified in
1963 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1964 systems where "make" and "gmake" are different, calling "gmake" at top
1965 level broke things. I've arranged for the value of $(MAKE) to be passed
1966 from the Makefile to this script so that it can call the same version of
1970 A note about Exim versions 4.44 and 4.50
1971 ----------------------------------------
1973 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1974 changes of various kinds. As a consequence, a big documentation update was
1975 needed. This delayed the release for rather longer than seemed good, especially
1976 in the light of a couple of (minor) security issues. Therefore, the changes
1977 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1978 release. So 4.44 and 4.50 are in effect two different branches that both start
1981 I have left the 4.50 change log unchanged; it contains all the changes since
1982 4.43. The change log for 4.44 is below; many of its items are identical to
1983 those for 4.50. This seems to be the most sensible way to preserve the
1984 historical information.
1990 1. Minor wording change to the doc/README.SIEVE file.
1992 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1993 computation of the current number of files was incorrect.
1995 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1996 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1999 4. Give more explanation in the error message when the command for a transport
2000 filter fails to execute.
2002 5. There are several places where Exim runs a non-Exim command in a
2003 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2004 was being done only for the command run by the queryprogram router. It is
2005 now done for all such subprocesses. The other cases are: ${run, transport
2006 filters, and the commands run by the lmtp and pipe transports.
2008 6. Added CONFIGURE_GROUP build-time option.
2010 7. Some older OS have a limit of 256 on the maximum number of file
2011 descriptors. Exim was using setrlimit() to set 1000 as a large value
2012 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2013 systems. I've change it so that if it can't get 1000, it tries for 256.
2015 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2016 was an oversight, and furthermore, ever since the addition of extra
2017 controls (e.g. 4.43/32), the checks on when to allow different forms of
2018 "control" were broken. There should now be diagnostics for all cases when a
2019 control that does not make sense is encountered.
2021 9. Added the /retain_sender option to "control=submission".
2023 10. $recipients is now available in the predata ACL (oversight).
2025 11. Tidy the search cache before the fork to do a delivery from a message
2026 received from the command line. Otherwise the child will trigger a lookup
2027 failure and thereby defer the delivery if it tries to use (for example) a
2028 cached ldap connection that the parent has called unbind on.
2030 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2031 of $address_data from the recipient verification was clobbered by the
2032 sender verification.
2034 13. The value of address_data from a sender verification is now available in
2035 $sender_address_data in subsequent conditions in the ACL statement.
2037 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2039 15. Added a new option "connect=<time>" to callout options, to set a different
2042 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2043 was its contents. (It was OK if the option was not defined at all.)
2045 17. A "Completed" log line is now written for messages that are removed from
2046 the spool by the -Mrm option.
2048 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2049 information about exactly what failed.
2051 19. Added -dd to debug only the daemon process.
2053 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2054 handles timeouts, both on the server side and network timeouts. Renamed the
2055 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2057 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2058 It is now set to "smtps".
2060 22. $host_address is now set to the target address during the checking of
2061 ignore_target_hosts.
2063 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2064 being passed; this would have caused $sender_host_name to have been used if
2065 matching the list had actually called for a host name (not very likely,
2066 since this list is usually IP addresses). A host name is now passed as
2069 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2070 code that specifies a non-restarting handler (typically sigaction() in
2071 modern systems) in an attempt to fix a rare and obscure crash bug.
2073 25. Narrowed the window for a race in the daemon that could cause it to ignore
2074 SIGCHLD signals. This is not a major problem, because they are used only to
2075 wake it up if nothing else does.
2077 26. A malformed maildirsize file could cause Exim to calculate negative values
2078 for the mailbox size or file count. Odd effects could occur as a result.
2079 The maildirsize information is now recalculated if the size or filecount
2082 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2083 support for a long time. Removed HAVE_SYS_VFS_H.
2085 28. Installed the latest version of exipick from John Jetmore.
2087 29. In an address list, if the pattern was not a regular expression, an empty
2088 subject address (from a bounce message) matched only if the pattern was an
2089 empty string. Non-empty patterns were not even tested. This was the wrong
2090 because it is perfectly reasonable to use an empty address as part of a
2091 database query. An empty address is now tested by patterns that are
2092 lookups. However, all the other forms of pattern expect the subject to
2093 contain a local part and a domain, and therefore, for them, an empty
2094 address still always fails if the pattern is not itself empty.
2096 30. Exim went into a mad DNS loop when attempting to do a callout where the
2097 host was specified on an smtp transport, and looking it up yielded more
2098 than one IP address.
2100 31. Re-factored the code for checking spool and log partition space into a
2101 function that finds that data and another that does the check. The former
2102 is then used to implement four new variables: $spool_space, $log_space,
2103 $spool_inodes, and $log_inodes.
2105 32. The RFC2047 encoding function was originally intended for short strings
2106 such as real names; it was not keeping to the 75-character limit for
2107 encoded words that the RFC imposes. It now respects the limit, and
2108 generates multiple encoded words if necessary. To be on the safe side, I
2109 have increased the buffer size for the ${rfc2047: expansion operator from
2112 33. It is now permitted to omit both strings after an "if" condition; if the
2113 condition is true, the result is "true". As before, when the second string
2114 is omitted, a false condition yields an empty string. This makes it less
2115 cumbersome to write custom ACL and router conditions.
2117 34. Failure to deliver a bounce message always caused it to be frozen, even if
2118 there was an errors_to setting on the router. The errors_to setting is now
2121 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2122 canonical form (fully expanded) before being placed in
2123 $sender_host_address.
2125 36. The table in the code that translates DNS record types into text (T_A to
2126 "A" for instance) was missing entries for NS and CNAME. It is just possible
2127 that this could have caused confusion if both these types were looked up
2128 for the same domain, because the text type is used as part of Exim's
2129 per-process caching. But the chance of anyone hitting this buglet seems
2132 37. The dnsdb lookup has been extended in a number of ways.
2134 (1) There is a new type, "zns", which walks up the domain tree until it
2135 finds some nameserver records. It should be used with care.
2137 (2) There is a new type, "mxh", which is like "mx" except that it returns
2138 just the host names, not the priorities.
2140 (3) It is now possible to give a list of domains (or IP addresses) to be
2141 looked up. The behaviour when one of the lookups defers can be
2142 controlled by a keyword.
2144 (4) It is now possible to specify the separator character for use when
2145 multiple records are returned.
2147 38. The dnslists ACL condition has been extended: it is now possible to supply
2148 a list of IP addresses and/or domains to be looked up in a particular DNS
2151 39. Added log_selector=+queue_time_overall.
2153 40. When running the queue in the test harness, wait just a tad after forking a
2154 delivery process, to get repeatability of debugging output.
2156 41. Include certificate and key file names in error message when GnuTLS fails
2157 to set them up, because the GnuTLS error message doesn't include the name
2158 of the failing file when there is a problem reading it.
2160 42. Allow both -bf and -bF in the same test run.
2162 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2164 44. The "Exiscan patch" is now merged into the mainline Exim source.
2166 45. Sometimes the final signoff response after QUIT could fail to get
2167 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2168 < 0 before doing a fflush(). This bug looks as though it goes back to the
2169 introduction of TLS in release 3.20, but "sometimes" must have been rare
2170 because the tests only now provoked it.
2172 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2173 (this can affect the format of dates).
2175 47. exim_tidydb, when checking for the continued existence of a message for
2176 which it has found a message-specific retry record, was not finding
2177 messages that were in split spool directories. Consequently, it was
2178 deleting retry records that should have stayed in existence.
2180 48. Steve fixed some bugs in eximstats.
2182 49. The SPA authentication driver was not abandoning authentication and moving
2183 on to the next authenticator when an expansion was forced to fail,
2184 contradicting the general specification for all authenticators. Instead it
2185 was generating a temporary error. It now behaves as specified.
2187 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2188 (the order specifies the preference for clients). The order is now AES256,
2189 AES128, 3DES, ARCFOUR128.
2191 51. Small patch to Sieve code - explicitly set From: when generating an
2194 52. Exim crashed if a remote delivery caused a very long error message to be
2195 recorded - for instance if somebody sent an entire SpamAssassin report back
2196 as a large number of 550 error lines. This bug was coincidentally fixed by
2197 increasing the size of one of Exim's internal buffers (big_buffer) that
2198 happened as part of the Exiscan merge. However, to be on the safe side, I
2199 have made the code more robust (and fixed the comments that describe what
2202 53. Now that there can be additional text after "Completed" in log lines (if
2203 the queue_time_overall log selector is set), a one-byte patch to exigrep
2204 was needed to allow it to recognize "Completed" as not the last thing in
2207 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2208 patch that reportedly fixes this has been added. I am not expert enough to
2209 create a test for it. This is what the patch creator wrote:
2211 "I found a little strange behaviour of ldap code when working with
2212 Windows 2003 AD Domain, where users was placed in more than one
2213 Organization Units. When I tried to give exim partial DN, the exit code
2214 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2215 But simultaneously result of request was absolutely normal ldap result,
2216 so I produce this patch..."
2218 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2219 REFERENCE, so I have modified the code to exclude the patch when that macro
2222 55. Some experimental protocols are using DNS PTR records for new purposes. The
2223 keys for these records are domain names, not reversed IP addresses. The
2224 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2225 leaves it alone. Component reversal etc. now happens only for IP addresses.
2228 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2230 57. Double the size of the debug message buffer (to 2048) so that more of very
2231 long debug lines gets shown.
2233 58. The exicyclog utility now does better if the number of log files to keep
2234 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2236 59. Two changes related to the smtp_active_hostname option:
2238 (1) $smtp_active_hostname is now available as a variable.
2239 (2) The default for smtp_banner uses $smtp_active_hostname instead
2240 of $primary_hostname.
2242 60. The host_aton() function is supposed to be passed a string that is known
2243 to be a valid IP address. However, in the case of IPv6 addresses, it was
2244 not checking this. This is a hostage to fortune. Exim now panics and dies
2245 if the condition is not met. A case was found where this could be provoked
2246 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2247 components; fortuitously, this particular loophole had already been fixed
2248 by change 4.50/55 above.
2250 If there are any other similar loopholes, the new check in host_aton()
2251 itself should stop them being exploited. The report I received stated that
2252 data on the command line could provoke the exploit when Exim was running as
2253 exim, but did not say which command line option was involved. All I could
2254 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2255 running as the user.
2258 61. There was a buffer overflow vulnerability in the SPA authentication code
2259 (which came originally from the Samba project). I have added a test to the
2260 spa_base64_to_bits() function which I hope fixes it.
2263 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2264 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2266 63. The daemon start-up calls getloadavg() while still root for those OS that
2267 need the first call to be done as root, but it missed one case: when
2268 deliver_queue_load_max is set with deliver_drop_privilege. This is
2269 necessary for the benefit of the queue runner, because there is no re-exec
2270 when deliver_drop_privilege is set.
2272 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2273 This has been fixed.
2275 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2276 was in use, was not putting the data itself into the right store pool;
2277 consequently, it could be overwritten for a subsequent message in the same
2278 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2281 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2283 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2284 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2285 the function were treating the return as a boolean value, which happened to
2286 work because 0=false and not-0=true, but is not correct code.
2288 68. The host_aton() function was not handling scoped IPv6 addresses (those
2289 with, for example, "%eth0" on the end) correctly.
2291 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2292 negated items (that is, ~something) in unsigned ints. Some compilers
2293 apparently mutter when there is no cast.
2295 70. If an address verification called from an ACL failed, and did not produce a
2296 user-specific message (i.e. there was only a "system" message), nothing was
2297 put in $acl_verify_message. In this situation, it now puts the system
2300 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2301 session; change 4.31/43 added the unwanted input to the log line - except
2302 that it did not do this in the start of session case. It now does.
2304 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2305 This is wrong and can cause the other end to generate a synchronization
2306 error if it is another Exim or anything else that does the synchronization
2307 check. A QUIT command is no longer sent after a timeout.
2309 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2310 during host lookups.
2312 74. The defer_ok option of callout verification was not working if it was used
2313 when verifying addresses in header lines, that is, for this case:
2315 verify = header_sender/callout=defer_ok
2317 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2318 those file descriptors could be used for SMTP connections. If anything
2319 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2320 could be sent to the SMTP client, causing chaos. The daemon now opens
2321 stdin, stdout, and stderr to /dev/null when it puts itself into the
2324 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2325 log by default. The user can override this with suitable Perl magic.
2327 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2328 the log message when discard triggers, was not working for the DATA ACL or
2329 for the non-SMTP ACL.
2331 78. Error message wording change in sieve.c.
2333 79. If smtp_accept_max_per_host was set, the number of connections could be
2334 restricted to fewer than expected, because the daemon was trying to set up
2335 a new connection before checking whether the processes handling previous
2336 connections had finished. The check for completed processes is now done
2337 earlier. On busy systems, this bug wouldn't be noticed because something
2338 else would have woken the daemon, and it would have reaped the completed
2341 80. If a message was submitted locally by a user whose login name contained one
2342 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2343 It caused a spool format error. I have fixed the spool reading code. A
2344 related problem was that the "from" clause in the Received: line became
2345 illegal because of the space(s). It is now covered by ${quote_local_part.
2347 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2350 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2352 83. Patch from Sieve maintainer to fix unterminated string problem in
2353 "vacation" handling.
2355 84. Some minor changes to the Linux configuration files to help with other
2356 OS variants using glibc.
2358 85. One more patch for Sieve to update vacation handling to latest spec.
2361 ----------------------------------------------------
2362 See the note above about the 4.44 and 4.50 releases.
2363 ----------------------------------------------------
2369 1. Change 4.43/35 introduced a bug that caused file counts to be
2370 incorrectly computed when quota_filecount was set in an appendfile
2373 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2374 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2377 3. Give more explanation in the error message when the command for a transport
2378 filter fails to execute.
2380 4. There are several places where Exim runs a non-Exim command in a
2381 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2382 was being done only for the command run by the queryprogram router. It is
2383 now done for all such subprocesses. The other cases are: ${run, transport
2384 filters, and the commands run by the lmtp and pipe transports.
2386 5. Some older OS have a limit of 256 on the maximum number of file
2387 descriptors. Exim was using setrlimit() to set 1000 as a large value
2388 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2389 systems. I've change it so that if it can't get 1000, it tries for 256.
2391 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2392 was an oversight, and furthermore, ever since the addition of extra
2393 controls (e.g. 4.43/32), the checks on when to allow different forms of
2394 "control" were broken. There should now be diagnostics for all cases when a
2395 control that does not make sense is encountered.
2397 7. $recipients is now available in the predata ACL (oversight).
2399 8. Tidy the search cache before the fork to do a delivery from a message
2400 received from the command line. Otherwise the child will trigger a lookup
2401 failure and thereby defer the delivery if it tries to use (for example) a
2402 cached ldap connection that the parent has called unbind on.
2404 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2405 of $address_data from the recipient verification was clobbered by the
2406 sender verification.
2408 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2409 was its contents. (It was OK if the option was not defined at all.)
2411 11. A "Completed" log line is now written for messages that are removed from
2412 the spool by the -Mrm option.
2414 12. $host_address is now set to the target address during the checking of
2415 ignore_target_hosts.
2417 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2418 being passed; this would have caused $sender_host_name to have been used if
2419 matching the list had actually called for a host name (not very likely,
2420 since this list is usually IP addresses). A host name is now passed as
2423 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2424 code that specifies a non-restarting handler (typically sigaction() in
2425 modern systems) in an attempt to fix a rare and obscure crash bug.
2427 15. Narrowed the window for a race in the daemon that could cause it to ignore
2428 SIGCHLD signals. This is not a major problem, because they are used only to
2429 wake it up if nothing else does.
2431 16. A malformed maildirsize file could cause Exim to calculate negative values
2432 for the mailbox size or file count. Odd effects could occur as a result.
2433 The maildirsize information is now recalculated if the size or filecount
2436 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2437 support for a long time. Removed HAVE_SYS_VFS_H.
2439 18. Updated exipick to current release from John Jetmore.
2441 19. Allow an empty sender to be matched against a lookup in an address list.
2442 Previously the only cases considered were a regular expression, or an
2445 20. Exim went into a mad DNS lookup loop when doing a callout where the
2446 host was specified on the transport, if the DNS lookup yielded more than
2449 21. The RFC2047 encoding function was originally intended for short strings
2450 such as real names; it was not keeping to the 75-character limit for
2451 encoded words that the RFC imposes. It now respects the limit, and
2452 generates multiple encoded words if necessary. To be on the safe side, I
2453 have increased the buffer size for the ${rfc2047: expansion operator from
2456 22. Failure to deliver a bounce message always caused it to be frozen, even if
2457 there was an errors_to setting on the router. The errors_to setting is now
2460 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2461 canonical form (fully expanded) before being placed in
2462 $sender_host_address.
2464 24. Updated eximstats to version 1.33
2466 25. Include certificate and key file names in error message when GnuTLS fails
2467 to set them up, because the GnuTLS error message doesn't include the name
2468 of the failing file when there is a problem reading it.
2470 26. Expand error message when OpenSSL has problems setting up cert/key files.
2473 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2474 (this can affect the format of dates).
2476 28. exim_tidydb, when checking for the continued existence of a message for
2477 which it has found a message-specific retry record, was not finding
2478 messages that were in split spool directories. Consequently, it was
2479 deleting retry records that should have stayed in existence.
2481 29. eximstats updated to version 1.35
2482 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2483 1.35 - bugfix such that pie charts by volume are generated correctly
2485 30. The SPA authentication driver was not abandoning authentication and moving
2486 on to the next authenticator when an expansion was forced to fail,
2487 contradicting the general specification for all authenticators. Instead it
2488 was generating a temporary error. It now behaves as specified.
2490 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2491 (the order specifies the preference for clients). The order is now AES256,
2492 AES128, 3DES, ARCFOUR128.
2494 31. Small patch to Sieve code - explicitly set From: when generating an
2497 32. Exim crashed if a remote delivery caused a very long error message to be
2498 recorded - for instance if somebody sent an entire SpamAssassin report back
2499 as a large number of 550 error lines. This bug was coincidentally fixed by
2500 increasing the size of one of Exim's internal buffers (big_buffer) that
2501 happened as part of the Exiscan merge. However, to be on the safe side, I
2502 have made the code more robust (and fixed the comments that describe what
2505 33. Some experimental protocols are using DNS PTR records for new purposes. The
2506 keys for these records are domain names, not reversed IP addresses. The
2507 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2508 leaves it alone. Component reversal etc. now happens only for IP addresses.
2511 34. The host_aton() function is supposed to be passed a string that is known
2512 to be a valid IP address. However, in the case of IPv6 addresses, it was
2513 not checking this. This is a hostage to fortune. Exim now panics and dies
2514 if the condition is not met. A case was found where this could be provoked
2515 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2516 components; fortuitously, this particular loophole had already been fixed
2517 by change 4.50/55 or 4.44/33 above.
2519 If there are any other similar loopholes, the new check in host_aton()
2520 itself should stop them being exploited. The report I received stated that
2521 data on the command line could provoke the exploit when Exim was running as
2522 exim, but did not say which command line option was involved. All I could
2523 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2524 running as the user.
2527 35. There was a buffer overflow vulnerability in the SPA authentication code
2528 (which came originally from the Samba project). I have added a test to the
2529 spa_base64_to_bits() function which I hope fixes it.
2532 36. The daemon start-up calls getloadavg() while still root for those OS that
2533 need the first call to be done as root, but it missed one case: when
2534 deliver_queue_load_max is set with deliver_drop_privilege. This is
2535 necessary for the benefit of the queue runner, because there is no re-exec
2536 when deliver_drop_privilege is set.
2538 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2539 was in use, was not putting the data itself into the right store pool;
2540 consequently, it could be overwritten for a subsequent message in the same
2541 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2544 38. Sometimes the final signoff response after QUIT could fail to get
2545 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2546 < 0 before doing a fflush(). This bug looks as though it goes back to the
2547 introduction of TLS in release 3.20, but "sometimes" must have been rare
2548 because the tests only now provoked it.
2554 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2555 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2556 Now it returns an integer. A number of calls were still expecting a T/F
2557 return. Fortuitously, in all cases, the tests worked in OK situations,
2558 which is the norm. However, things would have gone wrong on any write
2559 failures on the smtp file descriptor. This function is used when sending
2560 messages over SMTP and also when doing verify callouts.
2562 2. When Exim is called to do synchronous delivery of a locally submitted
2563 message (the -odf or -odi options), it no longer closes stderr before doing
2566 3. Implemented the mua_wrapper option.
2568 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2570 5. Implemented the functions header_remove(), header_testname(),
2571 header_add_at_position(), and receive_remove_recipient(), and exported them
2574 6. If an ACL "warn" statement specified the addition of headers, Exim already
2575 inserted X-ACL-Warn: at the start if there was no header name. However, it
2576 was not making this test for the second and subsequent header lines if
2577 there were newlines in the string. This meant that an invalid header could
2578 be inserted if Exim was badly configured.
2580 7. Allow an ACL "warn" statement to add header lines at the start or after all
2581 the Received: headers, as well as at the end.
2583 8. Added the rcpt_4xx retry error code.
2585 9. Added postmaster_mailfrom=xxx to callout verification option.
2587 10. Added mailfrom=xxxx to the callout verification option, for verify=
2590 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2591 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2593 12. Inserted some casts to stop certain compilers warning when using pointer
2594 differences as field lengths or precisions in printf-type calls (mostly
2595 affecting debugging statements).
2597 13. Added optional readline() support for -be (dynamically loaded).
2599 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2600 same clock tick as a message's arrival, so that its received time was the
2601 same as the "first fail" time on the retry record, and that message
2602 remained on the queue past the ultimate address timeout, every queue runner
2603 would try a delivery (because it was past the ultimate address timeout) but
2604 after another failure, the ultimate address timeout, which should have then
2605 bounced the address, did not kick in. This was a "< instead of <=" error;
2606 in most cases the first failure would have been in the next clock tick
2607 after the received time, and all would be well.
2609 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2610 being recognized when the domain list was tested by the match_domain
2611 condition in an expansion string.
2613 16. Added the ${str2b64: operator.
2615 17. Exim was always calling setrlimit() to set a large limit for the number of
2616 processes, without checking whether the existing limit was already
2617 adequate. (It did check for the limit on file descriptors.) Furthermore,
2618 errors from getrlimit() and setrlimit() were being ignored. Now they are
2619 logged to the main and panic logs, but Exim does carry on, to try to do its
2620 job under whatever limits there are.
2622 18. Imported PCRE 5.0.
2624 19. Trivial typo in log message " temporarily refused connection" (the leading
2627 20. If the log selector return_path_on_delivery was set and an address was
2628 redirected to /dev/null, the delivery process crashed because it assumed
2629 that a return path would always be set for a "successful" delivery. In this
2630 case, the whole delivery is bypassed as an optimization, and therefore no
2633 21. Internal re-arrangement: the function for sending a challenge and reading
2634 a response while authentication was assuming a zero-terminated challenge
2635 string. It's now changed to take a pointer and a length, to allow for
2636 binary data in such strings.
2638 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2640 23. Exim was not respecting finduser_retries when seeking the login of the
2641 uid under which it was called; it was always trying 10 times. (The default
2642 setting of finduser_retries is zero.) Also, it was sleeping after the final
2643 failure, which is pointless.
2645 24. Implemented tls_on_connect_ports.
2647 25. Implemented acl_smtp_predata.
2649 26. If the domain in control=submission is set empty, Exim assumes that the
2650 authenticated id is a complete email address when it generates From: or
2651 Sender: header lines.
2653 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2654 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2655 chown and chgrp in /bin and hostname in /usr/bin.
2657 28. Exim was keeping the "process log" file open after each use, just as it
2658 does for the main log. This opens the possibility of it remaining open for
2659 long periods when the USR1 signal hits a daemon. Occasional processlog
2660 errors were reported, that could have been caused by this. Anyway, it seems
2661 much more sensible not to leave this file open at all, so that is what now
2664 29. The long-running daemon process does not normally write to the log once it
2665 has entered its main loop, and it closes the log before doing so. This is
2666 so that log files can straightforwardly be renamed and moved. However,
2667 there are a couple of unusual error situations where the daemon does write
2668 log entries, and I had neglected to close the log afterwards.
2670 30. The text of an SMTP error response that was received during a remote
2671 delivery was being truncated at 512 bytes. This is too short for some of
2672 the long messages that one sometimes sees. I've increased the limit to
2675 31. It is now possible to make retry rules that apply only when a message has a
2676 specific sender, in particular, an empty sender.
2678 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2679 it possible to be selective about when SMTP synchronization is enforced.
2681 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2683 32. Implemented hosts_connection_nolog.
2685 33. Added an ACL for QUIT.
2687 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2690 35. Added mailbox_size and mailbox_filecount to appendfile.
2692 36. Added control = no_multiline_responses to ACLs.
2694 37. There was a bug in the logic of the code that waits for the clock to tick
2695 in the case where the clock went backwards by a substantial amount such
2696 that the microsecond fraction of "now" was more than the microsecond
2697 fraction of "then" (but the whole seconds number was less).
2699 38. Added support for the libradius Radius client library this is found on
2700 FreeBSD (previously only the radiusclient library was supported).
2706 1. When certain lookups returned multiple values in the form name=value, the
2707 quoting of the values was not always being done properly. Specifically:
2708 (a) If the value started with a double quote, but contained no whitespace,
2710 (b) If the value contained whitespace other than a space character (i.e.
2711 tabs or newlines or carriage returns) it was not quoted.
2712 This fix has been applied to the mysql and pgsql lookups by writing a
2713 separate quoting function and calling it from the lookup code. The fix
2714 should probably also be applied to nisplus, ibase and oracle lookups, but
2715 since I cannot test any of those, I have not disturbed their existing code.
2717 2. A hit in the callout cache for a specific address caused a log line with no
2718 reason for rejecting RCPT. Now it says "Previous (cached) callout
2719 verification failure".
2721 3. There was an off-by-one bug in the queryprogram router. An over-long
2722 return line was truncated at 256 instead of 255 characters, thereby
2723 overflowing its buffer with the terminating zero. As well as fixing this, I
2724 have increased the buffer size to 1024 (and made a note to document this).
2726 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2727 when Exim is waiting for an SMTP response from a remote server, Exim
2728 restarts its select() call on the socket, thereby resetting its timeout.
2729 This is not a problem when such interrupts are rare. Somebody set up a cron
2730 job to run exiwhat every 2 minutes, which is less than the normal select()
2731 timeout (5 or 10 minutes). This meant that the select() timeout never
2732 kicked in because it was always reset. I have fixed this by comparing the
2733 time when an interrupt arrives with the time at the start of the first call
2734 to select(). If more time than the timeout has elapsed, the interrupt is
2735 treated as a timeout.
2737 5. Some internal re-factoring in preparation for the addition of Sieve
2738 extensions (by MH). In particular, the "personal" test is moved to a
2739 separate function, and given an option for scanning Cc: and Bcc: (which is
2740 not set for Exim filters).
2742 6. When Exim created an email address using the login of the caller as the
2743 local part (e.g. when creating a From: or Sender: header line), it was not
2744 quoting the local part when it contained special characters such as @.
2746 7. Installed new OpenBSD configuration files.
2748 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2749 try to make them clearer.
2751 9. Callout options, other than the timeout value, were being ignored when
2752 verifying sender addresses in header lines. For example, when using
2754 verify = header_sender/callout=no_cache
2756 the cache was (incorrectly) being used.
2758 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2759 only the Cygwin environment.
2761 11. When return_path_on_delivery was set as a log selector, if different remote
2762 addresses in the same message used different return paths and parallel
2763 remote delivery occurred, the wrong values would sometimes be logged.
2764 (Whenever a remote delivery process finished, the return path value from
2765 the most recently started remote delivery process was logged.)
2767 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2768 lines when AUTH and/or TLS are in use. This is the "received protocol"
2769 field. Exim used to use "asmtp" for authenticated SMTP, without any
2770 indication (in the protocol name) for TLS use. Now it follows the RFC and
2771 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2772 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2773 names appear in log lines as well as in Received: header lines.
2775 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2776 extensions, and comparison tests, and to fix some bugs.
2778 14. Changes to the "personal" filter test:
2780 (1) The test was buggy in that it was just doing the equivalent of
2781 "contains" tests on header lines. For example, if a user's address was
2782 anne@some.where, the "personal" test would incorrectly be true for
2784 To: susanne@some.where
2786 This test is now done by extracting each address from the header in turn,
2787 and checking the entire address. Other tests that are part of "personal"
2788 are now done using regular expressions (for example, to check local parts
2789 of addresses in From: header lines).
2791 (2) The list of non-personal local parts in From: addresses has been
2792 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2793 taken from the Sieve specification recommendations.
2795 (3) If the message contains any header line starting with "List-" it is
2796 treated as non-personal.
2798 (4) The test for "circular" in the Subject: header line has been removed
2799 because it now seems ill-conceived.
2801 15. Minor typos in src/EDITME comments corrected.
2803 16. Installed latest exipick from John Jetmore.
2805 17. If headers_add on a router specified a text string that was too long for
2806 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2807 of string_sprintf() is now avoided.
2809 18. $message_body_size was not set (it was always zero) when running the DATA
2810 ACL and the local_scan() function.
2812 19. For the "mail" command in an Exim filter, no default was being set for
2813 the once_repeat time, causing a random time value to be used if "once" was
2814 specified. (If the value happened to be <= 0, no repeat happened.) The
2815 default is now 0s, meaning "never repeat". The "vacation" command was OK
2816 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2817 (I found it when inspecting the code).
2819 20. There is now an overall timeout for performing a callout verification. It
2820 defaults to 4 times the callout timeout, which applies to individual SMTP
2821 commands during the callout. The overall timeout applies when there is more
2822 than one host that can be tried. The timeout is checked before trying the
2823 next host. This prevents very long delays if there are a large number of
2824 hosts and all are timing out (e.g. when the network connections are timing
2825 out). The value of the overall timeout can be changed by specifying an
2826 additional sub-option for "callout", called "maxwait". For example:
2828 verify = sender/callout=5s,maxwait=20s
2830 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2831 to the end before writing, but this should make it even safer).
2833 22. Exim was forgetting that it had advertised PIPELINING for the second and
2834 subsequent messages on an SMTP connection. It was also not resetting its
2835 memory on STARTTLS and an internal HELO.
2837 23. When Exim logs an SMTP synchronization error within a session, it now
2838 records whether PIPELINING has been advertised or not.
2840 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2841 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2842 rather than long int.
2844 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2846 26. Added the never_mail option to autoreply.
2852 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2853 crash if the getsockname() call failed; this can happen if a connection is
2854 closed very soon after it is established. The problem was simply in the
2855 order in which certain operations were done, causing Exim to try to write
2856 to the SMTP stream before it had set up the file descriptor. The bug has
2857 been fixed by making things happen in the correct order.
2863 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2864 before the connection was closed, thus losing the rejection response.
2866 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2867 some early Solaris releases, but causes trouble in current releases where
2868 socklen_t is defined.
2870 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2873 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2876 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2877 that releasing the top of it at the end releases what was used for sub-
2878 expansions (unless the block got too big). However, discard this block if
2879 the first thing is a variable or header, so that we can use its block when
2880 it is dynamic (useful for very large $message_headers, for example).
2882 6. Lookups now cache *every* query, not just the most recent. A new, separate
2883 store pool is used for this. It can be recovered when all lookup caches are
2884 flushed. Lookups now release memory at the end of their result strings.
2885 This has involved some general refactoring of the lookup sources.
2887 7. Some code has been added to the store_xxx() functions to reduce the amount
2888 of flapping under certain conditions.
2890 8. log_incoming_interface used to affect only the <= reception log lines. Now
2891 it causes the local interface and port to be added to several more SMTP log
2892 lines, for example "SMTP connection from", and rejection lines.
2894 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2896 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2898 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2899 could be overwritten at the end of the current message (or the start of a
2900 new message if it was set in a HELO ACL). The value is now preserved for
2901 the duration of the SMTP connection.
2903 12. If a transport had a headers_rewrite setting, and a matching header line
2904 contained an unqualified address, that address was qualified, even if it
2905 did not match any rewriting rules. The underlying bug was that the values
2906 of the flags that permit the existence of unqualified sender and recipient
2907 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2908 non-local messages, and by -bnq for local messages) were not being
2909 preserved with the message after it was received.
2911 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2912 "next input=" as part of the text comprising the host identity instead of
2913 the correct text. The code was using the same buffer for two different
2914 strings. However, depending on which order the printing function evaluated
2915 its arguments, the bug did not always show up. Under Linux, for example, my
2916 test suite worked just fine.
2918 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2919 doesn't work with some older versions of Perl. It has been changed to "my",
2920 which in any case is probably the better facility to use.
2922 15. A really picky compiler found some instances of statements for creating
2923 error messages that either had too many or two few arguments for the format
2926 16. The size of the buffer for calls to the DNS resolver has been increased
2927 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2928 for addresses that have a lot of PTR records. This alleviates a problem; it
2929 does not fully solve it.
2931 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2932 buffer now truncates the list and logs the incident, which is the same
2933 action as happens when Exim is looking up a host name and its aliases.
2934 Previously in this situation something unpredictable would happen;
2935 sometimes it was "internal error: store_reset failed".
2937 18. If a server dropped the connection unexpectedly when an Exim client was
2938 using GnuTLS and trying to read a response, the client delivery process
2939 crashed while trying to generate an error log message.
2941 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2942 string, for example:
2944 warn message = H1: something\nH2: something
2946 the text was added as a single header line from Exim's point of view
2947 though it ended up OK in the delivered message. However, searching for the
2948 second and subsequent header lines using $h_h2: did not work. This has been
2949 fixed. Similarly, if a system filter added multiple headers in this way,
2950 the routers could not see them.
2952 20. Expanded the error message when iplsearch is called with an invalid key to
2953 suggest using net-iplsearch in a host list.
2955 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2956 ACLs are no longer actually imposed (and a message to that effect is
2959 22. If a "gecos" field in a passwd entry contained escaped characters, in
2960 particular, if it contained a \" sequence, Exim got it wrong when building
2961 a From: or a Sender: header from that name. A second bug also caused
2962 incorrect handling when an unquoted " was present following a character
2963 that needed quoting.
2965 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2966 was not being matched caselessly.
2968 24. Arranged for all hyphens in the exim.8 source to be escaped with
2971 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2972 properly. Recipient callout cache records were still being keyed to include
2973 the sender, even when use_sender was set false. This led to far more
2974 callouts that were necessary. The sender is no longer included in the key
2975 when use_sender is false.
2977 26. Added "control = submission" modifier to ACLs.
2979 27. Added the ${base62d: operator to decode base 62 numbers.
2981 28. dnsdb lookups can now access SRV records.
2983 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2984 the configuration file.
2986 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2987 (-v) mode. This makes the output for a verbose queue run more intelligible.
2989 31. Added a use_postmaster feature to recipient callouts.
2991 32. Added the $body_zerocount variable, containing the number of binary zero
2992 bytes in the message body.
2994 33. The time of last modification of the "new" subdirectory is now used as the
2995 "mailbox time last read" when there is a quota error for a maildir
2998 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3000 35. Added +ignore_unknown as a special item in host lists.
3002 36. Code for decoding IPv6 addresses in host lists is now included, even if
3003 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3004 address was recognized as an IP address, but was then not correctly decoded
3005 into binary, causing unexpected and incorrect effects when compared with
3012 1. Very minor rewording of debugging text in manualroute to say "list of
3013 hosts" instead of "hostlist".
3015 2. If verify=header_syntax was set, and a header line with an unqualified
3016 address (no domain) and a large number of spaces between the end of the
3017 name and the colon was received, the reception process suffered a buffer
3018 overflow, and (when I tested it) crashed. This was caused by some obsolete
3019 code that should have been removed. The fix is to remove it!
3021 3. When running in the test harness, delay a bit after writing a bounce
3022 message to get a bit more predictability in the log output.
3024 4. Added a call to search_tidyup() just before forking a reception process. In
3025 theory, someone could use a lookup in the expansion of smtp_accept_max_
3026 per_host which, without the tidyup, could leave open a database connection.
3028 5. Added the variables $recipient_data and $sender_data which get set from a
3029 lookup success in an ACL "recipients" or "senders" condition, or a router
3030 "senders" option, similar to $domain_data and $local_part_data.
3032 6. Moved the writing of debug_print from before to after the "senders" test
3035 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3036 problems for message scanning, either using a data ACL, or using
3037 local_scan() because the Received: header was not generated till after they
3038 were called (in order to set the time as the time of reception completion).
3039 I have revised the way this works. The header is now generated after the
3040 body is received, but before the ACL or local_scan() are called. After they
3041 are run, the timestamp in the header is updated.
3047 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3048 before starting a queue runner without re-exec. This happened only when
3049 deliver_drop_privilege was set or when the Exim user was set to root. The
3050 effect of the bug was that timeouts during subsequent deliveries caused
3051 crashes instead of being properly handled. The handler is now left at its
3052 default (and expected) setting.
3054 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3055 message, again when deliver_drop_privilege is set or Exim is run as root.
3056 The bug described in (1) was not present in this case, but the tidying up
3057 of the other signals was missing. I have made the two cases consistent.
3059 3. The ignore_target_hosts setting on a manualroute router was being ignored
3060 for hosts that were looked up using the /MX notation.
3062 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3065 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3066 operated on the sender address. After changing the $sender_address to <>
3067 for the sender address verify, Exim was re-instated it as the original
3068 (before rewriting) address, but remembering that it had rewritten it, so it
3069 wasn't rewriting it again. This bug also had the effect of breaking the
3070 sender address verification caching when the sender address was rewritten.
3072 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3073 This has been changed so that if the ip literal address matches
3074 ignore_target_hosts, the router declines.
3076 7. Added expansion conditions match_domain, match_address, and match_local_
3077 part (NOT match_host).
3079 8. The placeholder for the Received: header didn't have a length field set.
3081 9. Added code to Exim itself and to exim_lock to test for a specific race
3082 condition that could lead to file corruption when using MBX delivery. The
3083 issue is with the lockfile that is created in /tmp. If this file is removed
3084 after a process has opened it but before that process has acquired a lock,
3085 there is the potential for a second process to recreate the file and also
3086 acquire a lock. This could lead to two Exim processes writing to the file
3087 at the same time. The added code performs the same test as UW imapd; it
3088 checks after acquiring the lock that its file descriptor still refers to
3089 the same named file.
3091 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3092 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3095 11. Added the smtp_active_hostname option. If used, this will typically be made
3096 to depend on the incoming interface address. Because $interface_address is
3097 not set up until the daemon has forked a reception process, error responses
3098 that can happen earlier (such as "too many connections") no longer contain
3101 12. If an expansion in a condition on a "warn" statement fails because a lookup
3102 defers, the "warn" statement is abandoned, and the next ACL statement is
3103 processed. Previously this caused the whole ACL to be aborted.
3105 13. Added the iplsearch lookup type.
3107 14. Added ident_timeout as a log selector.
3109 15. Added tls_certificate_verified as a log selector.
3111 16. Added a global option tls_require_ciphers (compare the smtp transport
3112 option of the same name). This controls incoming TLS connections.
3114 17. I finally figured out how to make tls_require_ciphers do a similar thing
3115 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3116 before starting the TLS session.
3118 18. Tabs are now shown as \t in -bP output.
3120 19. If the log selector return_path_on_delivery was set, Exim crashed when
3121 bouncing a message because it had too many Received: header lines.
3123 20. If two routers both had headers_remove settings, and the first one included
3124 a superfluous trailing colon, the final name in the first list and the
3125 first name in the second list were incorrectly joined into one item (with a
3126 colon in the middle).
3132 1. Added -C and -D options to the exinext utility, mainly to make it easier
3133 to include in the automated testing, but these could be helpful when
3134 multiple configurations are in use.
3136 2. The exinext utility was not formatting the output nicely when there was
3137 an alternate port involved in the retry record key, nor when there was a
3138 message id as well (for retries that were specific to a specific message
3139 and a specific host). It was also confused by IPv6 addresses, because of
3140 the additional colons they contain. I have fixed the IPv4 problem, and
3141 patched it up to do a reasonable job for IPv6.
3143 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3144 delivery, the log line now contains "pipelined" if PIPELINING was used.
3146 4. An SMTP transport process used to panic and die if the bind() call to set
3147 an explicit outgoing interface failed. This has been changed; it is now
3148 treated in the same way as a connect() failure.
3150 5. A reference to $sender_host_name in the part of a conditional expansion
3151 that was being skipped was still causing a DNS lookup. This no longer
3154 6. The def: expansion condition was not recognizing references to header lines
3155 that used bh_ and bheader_.
3157 7. Added the _cache feature to named lists.
3159 8. The code for checking quota_filecount in the appendfile transport was
3160 allowing one more file than it should have been.
3162 9. For compatibility with Sendmail, the command line option
3170 and sets the incoming protocol and host name (for trusted callers). The
3171 host name and its colon can be omitted when only the protocol is to be set.
3172 Note the Exim already has two private options, -pd and -ps, that refer to
3173 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3174 "s", but I don't think that's a major issue.
3176 10. A number of refactoring changes to the code, none of which should affect
3179 (a) The number of logging options was getting close to filling up the
3180 32-bit word that was used as a bit map. I have split them into two classes:
3181 those that are passed in the argument to log_write(), and those that are
3182 only ever tested independently outside of that function. These are now in
3183 separate 32-bit words, so there is plenty of room for expansion again.
3184 There is no change in the user interface or the logging behaviour.
3186 (b) When building, for example, log lines, the code previously used a
3187 macro that called string_cat() twice, in order to add two strings. This is
3188 not really sufficiently general. Furthermore, there was one instance where
3189 it was actually wrong because one of the argument was used twice, and in
3190 one call a function was used. (As it happened, calling the function twice
3191 did not affect the overall behaviour.) The macro has been replaced by a
3192 function that can join an arbitrary number of extra strings onto a growing
3195 (c) The code for expansion conditions now uses a table and a binary chop
3196 instead of a serial search (which was left over from when there were very
3197 few conditions). Also, it now recognizes conditions like "pam" even when
3198 the relevant support is not compiled in: a suitably worded error message is
3199 given if an attempt is made to use such a condition.
3201 11. Added ${time_interval:xxxxx}.
3203 12. A bug was causing one of the ddress fields not to be passed back correctly
3204 from remote delivery subprocesses. The field in question was not being
3205 subsequently used, so this caused to problems in practice.
3207 13. Added new log selectors queue_time and deliver_time.
3209 14. Might have fixed a bug in maildirsizefile handling that threw up
3210 "unexpected character" debug warnings, and recalculated the data
3211 unnecessarily. In any case, I expanded the warning message to give more
3214 15. Added the message "Restricted characters in address" to the statements in
3215 the default ACL that block characters like @ and % in local parts.
3217 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3218 Three changes have been made:
3220 (a) There was a serious bug; a negative response to MAIL caused the whole
3221 recipient domain to be cached as invalid, thereby blocking all messages
3222 to all local parts at the same domain, from all senders. This bug has
3223 been fixed. The domain is no longer cached after a negative response to
3224 MAIL if the sender used is not empty.
3226 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3229 (c) A new callout option, "use_sender" has been added for people who want
3230 the modified behaviour.
3236 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3239 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3240 indeed breaks things for older releases.
3242 3. Added additional logging to the case where there is a problem reading data
3243 from a filter that is running in a subprocess using a pipe, in order to
3244 try to track down a specific problem.
3246 4. Testing facility fudge: when running in the test harness and attempting
3247 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3248 getting "No route to host". Convert this to a timeout.
3250 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3253 6. Some OS don't have socklen_t but use size_t instead. This affects the
3254 fifth argument of getsockopt() amongst other things. This is now
3255 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3256 can be set for individual OS. I have set it for SunOS5, OSF1, and
3257 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3258 some earlier ones do not.
3260 7. Change 4.30/15 was not doing the test caselessly.
3262 8. The standard form for an IPv6 address literal was being rejected by address
3263 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3264 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3265 this, as well as the form without the "IPv6" on the front (but only when
3266 address literals are enabled, of course).
3268 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3270 10. Exim crashed if a message with an empty sender address specified by -f
3271 encountered a router with an errors_to setting. This could be provoked only
3272 by a command such as
3276 where an empty string was supplied; "<>" did not hit this bug.
3278 11. Installed PCRE release 4.5.
3280 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3281 remained set. It is now erased.
3283 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3284 times from message ids (which are base 36 rather than the normal 62).
3286 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3287 were being counted as actual protocol errors, and logged if the log
3288 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3289 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3290 and DATA following a set of rejected RCPTs do not count as protocol errors.
3291 In other words, Exim assumes they were pipelined, though this may not
3292 actually be the case. Of course, in all cases the client gets an
3293 appropriate error code.
3295 15. If a lookup fails in an ACL condition, a message about the failure may
3296 be available; it is used if testing the ACL cannot continue, because most
3297 such messages specify what the cause of the deferral is. However, some
3298 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3299 that caused an old message to be retained and used if a later statement
3300 caused a defer, replacing the real cause of the deferral.
3302 16. If an IP address had so many PTR records that the DNS lookup buffer
3303 was not large enough to hold them, Exim could crash while trying to process
3304 the truncated data. It now detects and logs this case.
3306 17. Further to 4.21/58, another change has been made: if (and only if) the
3307 first line of a message (the first header line) ends with CRLF, a bare LF
3308 in a subsequent header line has a space inserted after it, so as not to
3309 terminate the header.
3311 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3312 unnecessarily, used atoi() instead of strtol(), and didn't check the
3313 termination when getting file sizes from file names by regex.
3315 19. Completely re-implemented the support for maildirsize files, in the light
3316 of a number of problems with the previous contributed implementation
3317 (4.30/29). In particular:
3319 . If the quota is zero, the maildirsize file is maintained, but no quota is
3322 . If the maildir directory does not exist, it is created before any attempt
3323 to write a maildirsize file.
3325 . The quota value in the file is just a cache; if the quota is changed in
3326 the transport, the new value overrides.
3328 . A regular expression is available for excluding directories from the
3331 20. The autoreply transport checks the characters in options that define the
3332 message's headers; it allows continued headers, but it was checking with
3333 isspace() after an embedded newline instead of explicitly looking for a
3336 21. If all the "regular" hosts to which an address was routed had passed their
3337 expiry times, and had not reached their retry times, the address was
3338 bounced, even if fallback hosts were defined. Now Exim should go on to try
3341 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3342 equivalent code in the SMTP transport. Some hosts send humungous responses
3343 to HELO/EHLO, more than 1024 it seems.
3345 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3346 gives compiler warnings in some environments. I've now done it "properly",
3349 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3350 (because of problems with the built-in one) was declared to return uschar *
3351 instead of char *, causing compiler failure.
3353 25. Fixed a file descriptor leak when processing alias/forward files.
3355 26. Fixed a minor format string issue in dbfn.c.
3357 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3359 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3360 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3363 29. When a sender address is verified, it is cached, to save repeating the test
3364 when there is more than one recipient in a message. However, when the
3365 verification involves a callout, it is possible for different callout
3366 options to be set for different recipients. It is too complicated to keep
3367 track of this in the cache, so now Exim always runs a verification when a
3368 callout is required, relying on the callout cache for the optimization.
3369 The overhead is duplication of the address routing, but this should not be
3372 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3373 to be verified with callout=postmaster, and the main callout worked but the
3374 postmaster check failed, the verification correctly failed. However, if a
3375 subsequent RCPT command asked for sender verification *without* the
3376 postmaster check, incorrect caching caused this verification also to fail,
3379 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3380 it was not caching the DNS options (qualify_single, search_parents) that
3381 were used when the lookup failed. A subsequent lookup with different
3382 options therefore always gave the same answer, though there were cases
3383 where it should not have. (Example: a "domains = !$mx_any" option on a
3384 dnslookup router: the "domains" option is always processed without any
3385 widening, but the router might have qualify_single set.) Now Exim uses the
3386 cached value only when the same options are set.
3388 32. Added John Jetmore's "exipick" utility to the distribution.
3390 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3391 than a timeout (e.g. a certificate is required, and is not provided), an
3392 Exim server now closes the connection immediately. Previously it waited for
3393 the client to close - but if the client is SSL, it seems that they each
3394 wait for each other, leading to a delay before one of them times out.
3396 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3397 maintained 0.8.x compatibility because I don't think many are using it, and
3398 it is clearly obsolete.
3400 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3403 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3404 client certificate was expired. A simple patch fixes this, though I don't
3405 understand the full logic of why the verify callback is called multiple
3408 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3409 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3410 which causes problems with some clients (such as the Certicom SSL Plus
3411 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3412 disables the coutermeasure allowing Eudora to connect."
3414 38. Exim was not checking that a write() to a log file succeeded. This could
3415 lead to Bad Things if a log got too big, in particular if it hit a file
3416 size limit. Exim now panics and dies if it cannot write to a log file, just
3417 as it does if it cannot open a log file.
3419 39. Modified OS/Makefile-Linux so that it now contains
3421 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3423 The two -D definitions ensure that Exim is compiled with large file
3424 support, which makes it possible to handle log files that are bigger than
3427 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3428 instance) a domain was checked against a named list that involved a lookup,
3429 causing $domain_data to be set, then another domain was checked against the
3430 same list, then the first domain was re-checked, the value of $domain_data
3431 after the final check could be wrong. In particular, if the second check
3432 failed, it could be set empty. This bug probably also applied to
3435 41. The strip_trailing_dot option was not being applied to the address given
3436 with the -f command-line option.
3438 42. The code for reading a message's header from the spool was incrementing
3439 $received_count, but never initializing it. This meant that the value was
3440 incorrect (doubled) while delivering a message in the same process in which
3441 it was received. In the most common configuration of Exim, this never
3442 happens - a fresh exec is done - but it can happen when
3443 deliver_drop_privilege is set.
3445 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3446 it now includes up to 150 characters of the unexpected data in the log
3449 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3450 and building data strings. The size of both of these buffers was 10 000
3451 bytes - far larger than anybody would *ever* want, thought I. Needless to
3452 say, somebody hit the limit. I have increased the maximum line length to
3453 20 000 and the maximum data length of concatenated lines to 100 000. I have
3454 also fixed two bugs, because there was no checking on these buffers. Tsk,
3455 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3456 buffer is too small.
3458 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3459 lsearch lookups. Now it does.
3461 46. When parsing a route_list item in a manualroute router, a fixed-length
3462 buffer was used for the list of hosts. I made this 1024 bytes long,
3463 thinking that nobody would ever have a list of hosts that long. Wrong.
3464 Somebody had a whole pile of complicated expansion conditions, and the
3465 string was silently truncated, leading to an expansion error. It turns out
3466 that it is easier to change to an unlimited length (owing to other changes
3467 that have happened since this code was originally written) than to build
3468 structure for giving a limitation error. The length of the item that
3469 expands into the list of hosts is now unlimited.
3471 47. The lsearch lookup could not handle data where the length of text line was
3472 more than 4095 characters. Such lines were truncated, leading to shortened
3473 data being returned. It should now handle lines of any length.
3475 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3476 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3479 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3480 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3481 no longer happens in comment lines. A list of replacements is now placed
3482 at the head of all of the source files, except those whose only change is
3483 to replace PERL_COMMAND in the very first #! line.
3485 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3486 messages on the queue, with a bottom-up merge sort, using code contributed
3487 by Michael Haardt. This should make operations like -bp somewhat faster on
3488 large queues. It won't affect queue runners, except when queue_run_in_order
3491 51. Installed eximstats 1.31 in the distribution.
3493 52. Added support for SRV lookups to the dnslookup router.
3495 53. If an ACL referred to $message_body or $message_body_end, the value was not
3496 reset for any messages that followed in the same SMTP session.
3498 54. The store-handling optimization for building very long strings was not
3499 differentiating between the different store pools. I don't think this
3500 actually made any difference in practice, but I've tidied it.
3502 55. While running the routers to verify a sender address, $sender_address
3503 was still set to the sender address. This is wrong, because when routing to
3504 send a bounce to the sender, it would be empty. Therefore, I have changed
3505 it so that, while verifying a sender address, $sender_address is set to <>.
3506 (There is no change to what happens when verifying a recipient address.)
3508 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3509 target A or AAAA records (if not already returned) without resetting the
3510 qualify_single or search_parents options of the DNS resolver. These are
3511 inappropriate in this case because the targets of MX and SRV records must
3512 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3513 target that, when qualified, matched something in the local domain. These
3514 two options are now turned off when doing these lookups.
3516 57. It seems that at least some releases of Reiserfs (which does not have the
3517 concept of a fixed number of inodes) returns zero and not -1 for the
3518 number of available inodes. This interacted badly with check_spool_inodes,
3519 which assumed that -1 was the "no such thing" setting. What I have done is
3520 to check that the total number of inodes is greater than zero before doing
3521 the test of how many are available.
3523 58. When a "warn" ACL statement has a log_message modifier, the message is
3524 remembered, and not repeated. This is to avoid a lot of repetition when a
3525 message has many recipients that cause the same warning to be written.
3526 Howewer, Exim was preserving the list of already written lines for an
3527 entire SMTP session, which doesn't seem right. The memory is now reset if a
3528 new message is started.
3530 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3531 debugging output unless log_rewrite was also set.
3533 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3534 of (int)(handle) when we know that handle contains (void *)(-1).
3536 61. The Exim daemon panic-logs an error return when it closes the incoming
3537 connection. However "connection reset by peer" seems to be common, and
3538 isn't really an error worthy of noting specially, so that particular error
3541 62. When Exim is trying to find all the local interfaces, it used to panic and
3542 die if the ioctl to get the interface flags failed. However, it seems that
3543 on at least one OS (Solaris 9) it is possible to have an interface that is
3544 included in the list of interfaces, but for which you get a failure error
3545 for this call. This happens when the interface is not "plumbed" into a
3546 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3547 failure of the "get flags" call assumes that the interface is down.
3549 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3550 makes life easier for people who are doing arithmetic on fields extracted
3551 from dates, where you often get leading zeros that should not be
3552 interpreted as octal.
3554 64. Added qualify_domain to the redirect router, to override the global
3557 65. If a pathologically long header line contained very many addresses (the
3558 report of this problem mentioned 10 000) and each of them was rewritten,
3559 Exim could use up a very large amount of memory. (It kept on making new
3560 copies of the header line as it rewrote, and never released the old ones.)
3561 At the expense of a bit more processing, the header rewriting function has
3562 been changed so that it no longer eats memory in this way.
3564 66. The generation of the Received: header has been moved from the time that a
3565 message starts to be received, to the time that it finishes. The timestamp
3566 in the Received: header should now be very close to that of the <= log
3567 line. There are two side-effects of this change:
3569 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3570 logged header lines no longer include the local Received: line, because
3571 it has not yet been created. The same applies to a copy of the message
3572 that is returned to a non-SMTP sender when a message is rejected.
3574 (b) When a filter file is tested using -bf, no additional Received: header
3575 is added to the test message. After some thought, I decided that this
3578 This change does not affect the value of $received_for. It is still set
3579 after address rewriting, but before local_scan() is called.
3581 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3583 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3584 gave an unhelpful panic error message, and a defer error. I have managed to
3585 change this behaviour so that it now rejects any supplied certificate,
3586 which seems right, as the list of acceptable certificates is empty.
3588 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3589 gave an unhelpful defer error. I have not managed to make this reject any
3590 supplied certificates, but the error message it gives is "no certificate
3591 supplied", which is not helpful.
3593 70. exigrep's output now also includes lines that are not associated with any
3594 message, but which match the given pattern. Implemented by a patch from
3595 Martin Sluka, which also tidied up the Perl a bit.
3597 71. Recipient callout verification, like sender verification, was using <> in
3598 the MAIL FROM command. This isn't really the right thing, since the actual
3599 sender may affect whether the remote host accepts the recipient or not. I
3600 have changed it to use the actual sender in the callout; this means that
3601 the cache record is now keyed on a recipient/sender pair, not just the
3602 recipient address. There doesn't seem to be a real danger of callout loops,
3603 since a callout by the remote host to check the sender would use <>.
3604 [SEE ABOVE: changed after hitting problems.]
3606 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3607 temporary errors. However, in the case of such a code being given after
3608 the end of a data transmission (i.e. after ".") Exim was failing to write
3609 a retry record for the message. (Yes, there was some broken host that was
3610 actually sending 8xx at this point.)
3612 73. An unknown lookup type in a host list could cause Exim to panic-die when
3613 the list was checked. (An example that provoked this was putting <; in the
3614 middle of a list instead of at the start.) If this happened during a DATA
3615 ACL check, a -D file could be left lying around. This kind of configuration
3616 error no longer causes Exim to die; instead it causes a defer errror. The
3617 incident is still logged to the main and panic logs.
3619 74. Buglet left over from Exim 3 conversion. The message "too many messages
3620 in one connection" was written to the rejectlog but not the mainlog, except
3621 when address rewriting (yes!) was being logged.
3623 75. Added write_rejectlog option.
3625 76. When a system filter was run not as root (that is, when system_filter_user
3626 was set), the values of the $n variables were not being returned to the
3627 main process; thus, they were not subsequently available in the $sn
3630 77. Added +return_path_on_delivery log selector.
3632 78. A connection timeout was being treated differently from recipients deferred
3633 when testing hosts_max_try with a message that was older than the host's
3634 retry timeout. (The host should not be counted, thus allowing all hosts to
3635 be tried at least once before bouncing.) This may have been the cause of an
3636 occasionally reported bug whereby a message would remain on the queue
3637 longer than the retry timeout, but would be bounced if a delivery was
3638 forced. I say "may" because I never totally pinned down the problem;
3639 setting up timeout/retry tests is difficult. See also the next item.
3641 79. The ultimate address timeout was not being applied to errors that involved
3642 a combination of host plus message (for example, a timeout on a MAIL
3643 command). When an address resolved to a number of possible hosts, and they
3644 were not all tried for each delivery (e.g. because of hosts_max_try), a
3645 message could remain on the queue longer than the retry timeout.
3647 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3650 81. Fixed an obscure SMTP outgoing bug which required at least the following
3651 conditions: (a) there was another message waiting for the same server;
3652 (b) the server returned 5xx to all RCPT commands in the first message so
3653 that the message was not completed; (c) the server dropped the connection
3654 or gave a negative response to the RSET that Exim sends to abort the
3655 transaction. The observed case was a dropped connection after DATA that had
3656 been sent in pipelining mode. That is, the server had advertised PIPELINING
3657 but was not implementing it correctly. The effect of the bug was incorrect
3658 behaviour, such as trying another host, and this could lead to a crash.
3664 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3665 and daemon.c were passed as pointers to ints; they should have been
3666 pointers to socklen_t variables (which are typically unsigned ints).
3668 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3671 3. Fixed a really odd bug that affected only the testing scheme; patching a
3672 certain fixed string in the binary changed the value of another string that
3673 happened to be identical to the end of the original first string.
3675 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3676 name", it returns that address as the IP address. On some operating
3677 systems (e.g. Solaris), it also passes back the IP address string as the
3678 "host name". However, on others (e.g. Linux), it passes back an empty
3679 string. Exim wasn't checking for this, and was changing the host name to an
3680 empty string, assuming it had been canonicized.
3682 5. Although rare, it is permitted to have more than one PTR record for a given
3683 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3684 all the names associated with an address, because they do in Solaris.
3685 However, it seems that they do not in Linux for data that comes from the
3686 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3687 I found this out when I moved to a new Linux workstation and tried to run
3688 the Exim test suite.
3690 To get round this problem I have changed the code so that it now does its
3691 own call to the DNS to look up PTR records when searching for a host name.
3692 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3693 addresses that are only in /etc/hosts are still found.
3695 This behaviour is, however, controlled by an option called host_lookup_
3696 order, which defaults to "bydns:byaddr". If people want to use the other
3697 order, or indeed, just use one or the other means of lookup, they can
3698 specify it in this variable.
3700 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3701 some operating systems, this comes back from gethostbyaddr() as an empty
3702 string, and this is what Exim used to test for. However, it seems that in
3703 other systems, "." is yielded. Exim now tests for this case too.
3705 7. The values of check_spool_space and check_log_space are now held internally
3706 as a number of kilobytes instead of an absolute number of bytes. If a
3707 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3708 kilobyte. This means that much larger values can be stored.
3710 8. Exim monitor: an attempt to get the action menu when not actually pointing
3711 at a message produces an empty menu entitled "No message selected". This
3712 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3713 no entries in it ("Shell widget menu has zero width and/or height"). So I
3714 have added a single, blank menu entry in this case.
3716 9. Added ${quote_local_part.
3718 10. MIME decoding is now applied to the contents of Subject: header lines when
3721 11. Now that a reference to $sender_host_address automatically causes a reverse
3722 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3723 host lookup before query-style lookups in lists that might use this
3724 variable. This has therefore been abolished, and the "net-" prefix is no
3725 longer necessary for query-style lookups.
3727 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3728 have been a typo for LFLAGS, so it has been changed.
3730 13. The install script calls Exim with "-C /dev/null" in order to find the
3731 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3732 to be output. Howeve, since Exim outputs its version number before the
3733 error, it didn't break the script. It just looked ugly. I fixed this by
3734 always allowing "-C /dev/null" if the caller is root.
3736 14. Ignore overlarge ACL variable number when reading spool file - insurance
3737 against a later release with more variables having written the file.
3739 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3740 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3741 this, as well as the form without the "IPv6" on the front.
3743 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3744 OS/Makefile-Darwin file.
3746 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3747 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3749 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3750 variable. However, some OS use ints for inodes. I've added cast to long int
3751 to get rid of the compiler warning.
3753 19. I had forgotten to lock out "/../" in configuration file names when
3754 ALT_CONFIG_PREFIX was set.
3756 20. Routers used for verification do not need to specify transports. However,
3757 if such a router generated a host list, and callout was configured, Exim
3758 crashed, because it could not find a port number from the (non-existent)
3759 transport. It now assumes port 25 in this circumstance.
3761 21. Added the -t option to exigrep.
3763 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3764 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3765 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3766 them is set, LOOKUP_LSEARCH is forced.
3768 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3770 24. Added sender and host information to the "rejected by local_scan()" log
3771 line; previously there was no indication of these.
3773 25. Added .include_if_exists.
3775 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3776 Linux. It turns out that not all file systems support this. Apparently some
3777 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3778 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3779 not supported on the file descriptor, is now ignored when Exim is trying to
3780 sync a directory. This applies only to Linux.
3782 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3784 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3785 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3788 29. Added support for maildirsize files from supplied patch (modified a bit).
3790 30. The use of :fail: followed by an empty string could lead Exim to respond to
3791 sender verification failures with (e.g.):
3793 550 Verification failed for <xxx>
3794 550 Sender verify failed
3796 where the first response line was missing the '-' that indicates it is not
3797 the final line of the response.
3799 31. The loop for finding the name of the user that called Exim had a hardwired
3800 limit of 10; it now uses the value of finduser_retries, which is used for
3801 all other user lookups.
3803 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3806 33. Exim was neglecting to zero errno before one call of strtol() when
3807 expanding a string and expecting an integer value. On some systems this
3808 resulted in spurious "integer overflow" errors. Also, it was casting the
3809 result into an int without checking.
3811 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3812 did not work. The code looks as if it has *never* worked, though it appears
3813 to have been documented since at least releast 1.62. I have made it work.
3815 35. The "timeout_DNS" error in retry rules, also documented since at least
3816 1.62, also never worked. As it isn't clear exactly what this means, and
3817 clearly it isn't a major issue, I have abolished the feature by treating it
3818 as "timeout", and writing a warning to the main and panic logs.
3820 36. The display of retry rules for -brt wasn't always showing the error code
3823 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3824 timeout_connect_A, timeout_connect_MX.
3826 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3827 to the empty sender.
3829 39. The daemon was not analysing the content of -oX till after it had closed
3830 stderr and disconnected from the controlling terminal. This meant that any
3831 syntax errors were only noted on the panic log, and the return code from
3832 the command was 0. By re-arranging the code a little, I've made the
3833 decoding happen first, so such errors now appear on stderr, and the return
3834 code is 1. However, the actual setting up of the sockets still happens in
3835 the disconnected process, so errors there are still only recorded on the
3838 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3839 connections (as happens on some IP stacks) was logged at start up time as
3840 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3841 it from "IPv6 and IPv4", which means that two separate sockets are being
3844 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3845 says whether AF_INET or AF_INET6 was passed as an argument.
3847 42. Exiwhat output was messed up when time zones were included in log
3850 43. Exiwhat now gives more information about the daemon's listening ports,
3851 and whether -tls-on-connect was used.
3853 44. The "port" option of the smtp transport is now expanded.
3855 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3856 silently ignored. Now an error message is written to the main and panic
3859 46. There's a new ACL modifier called "logwrite" which writes to a log file
3860 as soon as it is encountered.
3862 47. Added $local_user_uid and $local_user_gid at routing time.
3864 48. Exim crashed when trying to verify a sender address that was being
3867 49. Exim was recognizing only a space character after ".include". It now also
3868 recognizes a tab character.
3870 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3871 extracting the relevant information from the specification. The man page no
3872 longer contains scrambled data for the -d option, and I've added a section
3873 at the front about calling Exim under different names.
3875 51. Added "extra_headers" argument to the "mail" command in filter files.
3877 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3880 53. Installed eximstats 1.29.
3882 54. Added transport_filter_timeout as a generic transport option.
3884 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3885 Cc: header lines. This was required by RFC 822, but it not required by RFC
3888 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3889 incoming messages that did not have them. Now it does so only if the
3890 message originates locally, that is, if there is no associated remote host
3891 address. When Resent- header lines are present, this applies to the Resent-
3892 lines rather than the non-Resent- lines.
3894 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3895 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3897 58. Messages for configuration errors now include the name of the main
3898 configuration file - useful now that there may be more than one file in a
3899 list (.included file names were always shown).
3901 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3902 for those rare installations that do not start the daemon as root or run it
3903 setuid root. I've cut out the call to initgroups() if the daemon is not
3906 60. The Exim user and group can now be bound into the binary as text strings
3907 that are looked up at the start of Exim's processing.
3909 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3911 62. Added $mailstore_basename variable.
3913 63. Installed patch to sieve.c from Michael Haardt.
3915 64. When Exim failed to open the panic log after failing to open the main log,
3916 the original message it was trying to log was written to stderr and debug
3917 output, but if they were not available (the usual case in production), it
3918 was lost. Now it is written to syslog before the two lines that record the
3919 failures to open the logs.
3921 65. Users' Exim filters run in subprocesses under the user's uid. It is
3922 possible for a "deliver" command or an alias in a "personal" command to
3923 provoke an address rewrite. If logging of address rewriting is configured,
3924 this fails because the process is not running as root or exim. There may be
3925 a better way of dealing with this, but for the moment (because 4.30 needs
3926 to be released), I have disabled address rewrite logging when running a
3927 filter in a non-root, non-exim process.
3933 1. The buildconfig auxiliary program wasn't quoting the value set for
3934 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3935 not defined. This bug was masked in 4.22 by the effect that was fixed in
3938 2. Some messages that were rejected after a message id was allocated were
3939 shown as "incomplete" by exigrep. It no longer does this for messages that
3940 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3942 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3943 have allow_domain_literals set, the ID did not get logged in the <= line.
3944 Domain literals are now always recognized in Message-ID: header lines.
3946 4. The first argument for a ${extract expansion item is the key name or field
3947 number. Leading and trailing spaces in this item were not being ignored,
3948 causing some misleading effects.
3950 5. When deliver_drop_privilege was set, single queue runner processes started
3951 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3952 same command in the process it spins off) were not dropping privilege.
3954 6. When the daemon running as "exim" started a queue runner, it always
3955 re-executed Exim in the spun-off process. This is a waste of effort when
3956 deliver_drop_privilege is set. The new process now just calls the
3957 queue-runner function directly.
3963 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3966 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3967 ignored. Though the use of -oP was forcing the writing of a pid file, it
3968 was always written to the default place.
3970 3. If the message "no IP address found for host xxxx" is generated during
3971 incoming verification, it is now followed by identification of the incoming
3972 connection (so you can more easily find what provoked it).
3974 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3976 5. Added some features to "harden" Exim a bit more against certain attacks:
3978 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3979 be put in Local/Makefile. This is like the never_users runtime option,
3980 but it cannot be overridden. The default setting is "root".
3982 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3983 prefix string with which any file named in a -C command line option
3986 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3987 is retained for -C and -D only if the caller of Exim is root. Without
3988 it, the exim user may also use -C and -D and retain privilege.
3990 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3991 command line option is disabled.
3993 6. Macro names set by the -D option must start with an upper case letter, just
3994 like macro names defined in the configuration file.
3996 7. Added "dereference=" facility to LDAP.
3998 8. Two instances of the typo "uknown" in the source files are fixed.
4000 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4001 the Configure-Makefile script screwed up while processing it.
4003 10. Incorporated PCRE 4.4.
4005 11. The SMTP synchronization check was not operating right at the start of an
4006 SMTP session. For example, it could not catch a HELO sent before the client
4007 waited for the greeting. There is now a check for outstanding input at the
4008 point when the greeting is written. Because of the duplex, asynchronous
4009 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4010 way, but not yet received, when the check is performed.
4012 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4013 on TCP/IP sockets, because this apparently causes some broken clients to
4016 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4017 unchanged) from the Cygwin maintainer.
4019 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4020 was supported instead of testing for mbx. Effectively a typo.
4022 15. The spa authenticator server code was not checking that the input it
4023 received was valid base64.
4025 16. The debug output line for the "set" modifier in ACLs was not showing the
4026 name of the variable that was being set.
4028 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4030 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4031 lookup on its own. Something else was needed to trigger the lookup. For
4032 example, a match in host_lookup or the need for a host name in a host list.
4033 Now, if $sender_host_name is referenced and the host name has not yet been
4034 looked up, a lookup is performed. If the lookup fails, the variable remains
4035 empty, and $host_lookup_failed is set to "1".
4037 19. Added "eqi" as a case-independent comparison operator.
4039 20. The saslauthd authentication condition could segfault if neither service
4040 nor realm was specified.
4042 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4043 error message that was logged was misleading, and incoming SMTP
4044 connections were dropped. The message is now more accurate, and temporary
4045 errors are given to SMTP connections.
4047 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4048 (except RSET) with a 421 error, until QUIT is received. However, it was
4049 failing to send a response to QUIT.
4051 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4052 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4053 ACL is now run after the helo_try_verify_hosts code.
4055 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4056 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4057 case-independent, so other case variants are also recognized). Apparently
4058 some systems use these upper case variants.
4060 25. If more than two messages were waiting for the same host, and a transport
4061 filter was specified for the transport, Exim sent two messages over the
4062 same TCP/IP connection, and then failed with "socket operation on non-
4063 socket" when it tried to send the third.
4065 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4067 27. The extern definition of crypt16() in expand.c was not being excluded when
4068 the OS had its own crypt16() function.
4070 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4071 as a preferred synonym for return_size_limit, both as an option and as an
4074 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4076 30. Changed the default configuration ACL to relax the local part checking rule
4077 for addresses that are not in any local domains. For these addresses,
4078 slashes and pipe symbols are allowed within local parts, but the sequence
4079 /../ is explicitly forbidden.
4081 31. SPA server authentication was not clearing the challenge buffer before
4084 32. log_message in a "warn" ACL statement was writing to the reject log as
4085 well as to the main log, which contradicts the documentation and doesn't
4086 seem right (because no rejection is happening). So I have stopped it.
4088 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4089 However, I am unable to do any testing of this.
4091 34. Fixed an infelicity in the appendfile transport. When checking directories
4092 for a mailbox, to see if any needed to be created, it was accidentally
4093 using path names with one or more superfluous leading slashes; tracing
4094 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4096 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4097 contents are added to the log line that is written for every discarded
4098 recipient. (Previously a log_message setting was ignored.)
4100 36. The ${quote: operator now quotes the string if it is empty.
4102 37. The install script runs exim in order to find its version number. If for
4103 some reason other than non-existence or emptiness, which it checks, it
4104 could not run './exim', it was installing it with an empty version number,
4105 i.e. as "exim-". This error state is now caught, and the installation is
4108 38. An argument was missing from the function that creates an error message
4109 when Exim fails to connect to the socket for saslauthd authentication.
4110 This could cause Exim to crash, or give a corrupted message.
4112 39. Added isip, isip4, and isip6 to ${if conditions.
4114 40. The ACL variables $acl_xx are now saved with the message, and can be
4115 accessed later in routers, transports, and filters.
4117 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4118 strings in the file are not string-expanded.
4120 42. If a MAIL command specified a SIZE value that was too large to fit into an
4121 int variable, the check against message_size_limit failed. Such values are
4122 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4123 day this will have to be increased, but I don't think I want to be around
4124 when emails are that large.
4131 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4132 iconv() is not standard in FreeBSD.
4134 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4135 IPv6 enabled. The observed symptom was a segmentation fault on return from
4136 the function os_common_find_running_interfaces() in src/os.c.
4138 3. In the check_special_case() function in daemon.c I had used "errno" as an
4139 argument name, which causes warnings on some systems. This was basically a
4140 typo, since it was named "eno" in the comments!
4142 4. The code that waits for the clock to tick (at a resolution of some fraction
4143 of a second) so as to ensure message-id uniqueness was always waiting for
4144 at least one whole tick, when it could have waited for less. [This is
4145 almost certainly not relevant at current processor speeds, where it is
4146 unlikely to ever wait at all. But we try to future-proof.]
4148 5. The function that sleeps for a time interval that includes fractions of a
4149 second contained a race. It did not block SIGALRM between setting the
4150 timer, and suspending (a couple of lines later). If the interval was short
4151 and the sigsuspend() was delayed until after it had expired, the suspension
4152 never ended. On busy systems this could lead to processes getting stuck for
4155 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4156 process, before it forks any delivery processes. The open lookup caching
4157 mechanism meant that the open file or database connection was passed into
4158 the delivery process. The problem was that delivery processes always tidy
4159 up cached lookup data. This could cause a problem for the next delivery
4160 process started by the queue runner, because the external queue runner
4161 process does not know about the closure. So the next delivery process
4162 still has data in the lookup cache. In the case of a file lookup, there was
4163 no problem because closing a file descriptor in a subprocess doesn't affect
4164 the parent. However, if the lookup was caching a connection to a database,
4165 the connection was closed, and the second delivery process was likely to
4166 see errors such as "PGSQL: query failed: server closed the connection
4167 unexpectedly". The problem has been fixed by closing all cached lookups
4168 in a queue runner before running a delivery process.
4170 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4171 seem to have the "const" qualifier which it has on other OS. I've
4174 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4175 *of the same type* (client or server) with the same public name that an
4176 error should be diagnosed.
4178 9. When Exim looked up a host name for an IP address, but failed to find the
4179 original IP address when looking up the host name (a safety check), it
4180 output the message "<ip address> does not match any IP for NULL", which was
4181 confusing, to say the least. The bug was that the host name should have
4182 appeared instead of "NULL".
4184 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4185 user that is built into the binary, and the -C or -D options is used, root
4186 privilege is dropped before the configuration file is read. In addition,
4187 logging is switched to stderr instead of the normal log files. If the
4188 configuration then re-defines the Exim user, the unprivileged environment
4189 is probably not what is expected, so Exim logs a panic warning message (but
4192 However, if deliver_drop_privilege is set, the unprivileged state may well
4193 be exactly what is intended, so the warning has been cut out in that case,
4194 and Exim is allowed to try to write to its normal log files.
4200 1. smtp_return_error_details was not giving details for temporary sender
4201 or receiver verification errors.
4203 2. Diagnose a configuration error if two authenticators have the same public
4206 3. Exim used not to create the message log file for a message until the first
4207 delivery attempt. This could be confusing when incoming messages were held
4208 for policy or load reasons. The message log file is now created at the time
4209 the message is received, and an initial "Received" line is written to it.
4211 4. The automatically generated man page for command line options had a minor
4212 bug that caused no ill effects; however, a more serious problem was that
4213 the procedure for building the man page automatically didn't always
4214 operate. Consequently, release 4.20 contains an out-of-date version. This
4215 shouldn't happen again.
4217 5. When building Exim with embedded Perl support, the script that builds the
4218 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4219 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4221 6. The freeze_tell option was not being used for messages that were frozen on
4222 arrival, either by an ACL or by local_scan().
4224 7. Added the smtp_incomplete_transaction log selector.
4226 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4227 was accepting AUTH without a new EHLO.
4229 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4230 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4233 10. Logging of TCP/IP connections (when configured) now happens in the main
4234 daemon process instead of the child process, so that the TCP/IP connection
4235 count is more accurate (but it can never be perfect).
4237 11. The use of "drop" in a nested ACL was not being handled correctly in the
4238 outer ACL. Now, if condition failure induced by the nested "drop" causes
4239 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4240 or "require"), the connection is dropped.
4242 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4243 that yield "discard" can now be used with an "accept" or a "discard" verb,
4244 but an error is generated for any others (because I can't see a useful way
4245 to define what should happen).
4247 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4248 are now processed in the same way as lines in the Exim configuration file.
4249 In particular, continuation lines are supported.
4251 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4253 15. Added -ti meaning -t -i.
4255 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4256 domains, and warn by logging if others are found.
4258 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4259 returned from GIFCONF when Exim is trying to find the list of interfaces on
4260 a host. The code in os.c has been modified to copy each ifreq to an aligned
4261 structure in all cases.
4263 Also, in some cases, the returned ifreq's were being copied to a 'struct
4264 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4265 means the last couple of bytes of an IPv6 address could be chopped if the
4266 ifreq contained only a normal sockaddr (14 bytes storage).
4268 18. Named domain lists were not supported in the hosts_treat_as_local option.
4269 An entry such as +xxxx was not recognized, and was treated as a literal
4272 19. Ensure that header lines added by a DATA ACL are included in the reject log
4273 if the ACL subsequently rejects the message.
4275 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4276 MD5 (which is deprecated).
4278 21. When testing a filter file using -bf, Exim was writing a message when it
4279 took the sender from a "From " line in the message, but it was not doing so
4280 when it took $return_path from a Return-Path: header line. It now does.
4282 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4283 with a valid header line field name (a series of printing characters
4284 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4286 23. Changed "disc" in the source to "disk" to conform to the documentation and
4287 the book and for uniformity.
4289 24. Ignore Sendmail's -Ooption=value command line item.
4291 25. When execve() failed while trying to run a command in a pipe transport,
4292 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4293 could be confused with a return value of 69 from the command itself. This
4294 has been changed to 127, the value the shell returns if it is asked to run
4295 a non-existent command. The wording for the related log line suggests a
4296 non-existent command as the problem.
4298 26. If received_header_text expands to an empty string, do not add a Received:
4299 header line to the message. (Well, it adds a token one on the spool, but
4300 marks it "old" so that it doesn't get used or transmitted.)
4302 27. Installed eximstats 1.28 (addition of -nt option).
4304 28. There was no check for failure on the call to getsockname() in the daemon
4305 code. This can fail if there is a shortage of resources on the system, with
4306 ENOMEM, for example. A temporary error is now given on failure.
4308 29. Contrary to the C standard, it seems that in some environments, the
4309 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4310 program. Exim now does this explicitly; it affects the formatting of
4311 timestamps using strftime().
4313 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4314 complaints. I've now initialized all the variables, to avoid this.
4316 32. Header lines added by a system filter were not being "seen" during
4317 transport-time rewrites.
4319 33. The info_callback() function passed to OpenSSL is set up with type void
4320 (*)(SSL *, int, int), as described somewhere. However, when calling the
4321 function (actually a macro) that sets it up, the type void(*)() is
4322 expected. I've put in a cast to prevent warnings from picky compilers.
4324 34. If a DNS black list lookup found a CNAME record, but there were no A
4325 records associated with the domain it pointed at, Exim crashed.
4327 35. If a DNS black list lookup returned more than one A record, Exim ignored
4328 all but the first. It now scans all returned addresses if a particular IP
4329 value is being sought. In this situation, the contents of the
4330 $dnslist_value variable are a list of all the addresses, separated by a
4333 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4334 to accept a host name and all its aliases if the forward lookup for any of
4335 them yielded the IP address of the incoming connection. Now it accepts only
4336 those names whose forward lookup yields the correct IP address. Any other
4337 names are discarded. This closes a loophole whereby a rogue DNS
4338 administrator could create reverse DNS records to break through a
4339 wildcarded host restriction in an ACL.
4341 37. If a user filter or a system filter that ran in a subprocess used any of
4342 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4343 the wrong values were passed to the pipe command ($thisaddress had the
4344 value of $0, $0 had the value of $1, etc). This bug was introduced by
4345 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4347 38. Improved the line breaking for long SMTP error messages from ACLs.
4348 Previously, if there was no break point between 40 and 75 characters, Exim
4349 left the rest of the message alone. Two changes have been made: (a) I've
4350 reduced the minimum length to 35 characters; (b) if it can't find a break
4351 point between 35 and 75 characters, it looks ahead and uses the first one
4352 that it finds. This may give the occasional overlong line, but at least the
4353 remaining text gets split now.
4355 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4356 file descriptors might be low, and that setting 1000 would always raise it.
4357 It turns out that in some environments, the limit is already over 1000 and
4358 that lowering it causes trouble. So now Exim takes care not to decrease it.
4360 40. When delivering a message, the value of $return_path is set to $sender_
4361 address at the start of routing (routers may change the value). By an
4362 oversight, this default was not being set up when an address was tested by
4363 -bt or -bv, which affected the outcome if any router or filter referred to
4366 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4367 log only when "message" or "log_message" are set. However, if one of the
4368 conditions was an address verification, or a call to a nested ACL, the
4369 messages generated by the underlying test were being passed through. This
4370 no longer happens. The underlying message is available in $acl_verify_
4371 message for both "message" and "log_message" expansions, so it can be
4372 passed through if needed.
4374 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4375 new expansion $bh_ to give the encoded byte string without charset
4376 translation. Translation happens only if iconv() is available; HAVE_ICONV
4377 indicates this at build time. HEADERS_CHARSET gives the charset to
4378 translate to; headers_charset can change it in the configuration, and
4379 "headers charset" can change it in an individual filter file.
4381 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4382 that creates RFC 2047 encoded "words" labels them as that charset instead
4383 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4384 expansion operator; (ii) when Exim creates a From: line for a local
4385 message; (iii) when a header line is rewritten to include a "phrase" part.
4387 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4388 buggy, causing it to skip the first lines of messages whose message ID
4389 ended in 'D'. This would not have bitten before Exim release 4.14, because
4390 message IDs were unlikely to end in 'D' before then. The effect was to have
4391 incorrect size information for certain domains.
4393 45. #include "config.h" was missing at the start of the crypt16.c module. This
4394 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4397 46. If there was a timeout during a "random" callout check, Exim treated it as
4398 a failure of the random address, and carried on sending RSET and the real
4399 address. If the delay was just some slowness somewhere, the response to the
4400 original RCPT would be taken as a response to RSET and so on, causing
4401 mayhem of various kinds.
4403 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4404 when I implemented it. It didn't allow for the fact that some option values
4405 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4406 the right test for positive values.
4408 48. Domain names in DNS records are case-independent. Exim always looks them up
4409 in lower case. Some resolvers return domain names in exactly the case they
4410 appear in the zone file, that is, they may contain uppercase letters. Not
4411 all resolvers do this - some return always lower case. Exim was treating a
4412 change of case by a resolver as a change of domain, similar to a widening
4413 of a domain abbreviation. This triggered its re-routing code and so it was
4414 trying to route what was effectively the same domain again. This normally
4415 caused routing to fail (because the router wouldn't handle the domain
4416 twice). Now Exim checks for this case specially, and just changes the
4417 casing of the domain that it ultimately uses when it transmits the message
4420 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4423 50. If a filter generated a file delivery with a non-absolute name (possible if
4424 no home directory exists for the router), the forbid_file option was not
4427 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4428 the existing equality matching.
4430 52. Exim was using ints instead of ino_t variables in some places where it was
4431 dealing with inode numbers.
4433 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4434 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4435 called TMPDIR, and if it finds it is different, it changes its value.
4437 54. The smtp_printf() function is now made available to local_scan() so
4438 additional output lines can be written before returning. There is also an
4439 smtp_fflush() function to enable the detection of a dropped connection.
4440 The variables smtp_input and smtp_batched_input are exported to
4443 55. Changed the default runtime configuration: the message "Unknown user"
4444 has been removed from the ACL, and instead placed on the localuser router,
4445 using the cannot_route_message feature. This means that any verification
4446 failures that generate their own messages won't get overridden. Similarly,
4447 the "Unrouteable address" message that was in the ACL for unverifiable
4448 relay addresses has also been removed.
4450 56. Added hosts_avoid_esmtp to the smtp transport.
4452 57. The exicyclog script was not checking for the esoteric option
4453 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4454 will work only if exicyclog is run under the appropriate euid.
4456 58. Following a discussion on the list, the rules by which Exim recognises line
4457 endings on incoming messages have been changed. The -dropcr and drop_cr
4458 options are now no-ops, retained only for backwards compatibility. The
4459 following line terminators are recognized: LF CRLF CR. However, special
4460 processing applies to CR:
4462 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4463 nor a local message in the state where . is a terminator.
4465 (ii) If a bare CR is encountered in a header line, an extra space is added
4466 after the line terminator so as not to end the header. The reasoning
4467 behind this is that bare CRs in header lines are most likely either
4468 to be mistakes, or people trying to play silly games.
4470 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4471 was being incorrectly given as 18 bytes larger than it should have been.
4472 This is a VOB (very old bug).
4474 60. This may never have affected anything current, but just in case it has:
4475 When the local host is found other than at the start of a list of hosts,
4476 the local host, those with the same MX, and any that follow, are discarded.
4477 When the list in question was part of a longer list of hosts, the following
4478 hosts (not currently being processed) were also being discarded. This no
4479 longer happens. I'm not sure if this situation could ever has previously
4482 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4485 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4486 header. This is something that is recommended in a new Internet Draft, and
4487 is something that is documented as being done by Sendmail. There are two
4488 possible values. For messages generated by the autoreply transport, Exim
4491 Auto-Submitted: auto-replied
4493 whereas for all other generated messages (e.g. bounces) it adds
4495 Auto-Submitted: auto-generated
4497 63. The "personal" condition in filters now includes a test for the
4498 Auto-Submitted: header. If it contains the string "auto-" the message it
4499 not considered personal.
4501 64. Added rcpt_include_affixes as a generic transport option.
4503 65. Added queue_only_override (default true).
4505 66. Added the syslog_duplication option.
4507 67. If what should have been the first header line of a message consisted of
4508 a space followed by a colon, Exim was mis-interpreting it as a header line.
4509 It isn't of course - it is syntactically invalid and should therefore be
4510 treated as the start of the message body. The misbehaviour could have
4511 caused a number of strange effects, including loss of data in subsequent
4512 header lines, and spool format errors.
4514 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4515 client host had authenticated. This control can now be exercised by an ACL
4516 for more flexibility.
4518 69. By default, callouts do not happen when testing with -bh. There is now a
4519 variant, -bhc, which does actually run the callout code, including
4520 consulting and updating the callout cache.
4522 70. Added support for saslauthd authentication, courtesy of Alexander
4525 71. If statvfs() failed on the spool or log directories while checking their
4526 size for availability, Exim confusingly gave the error "space shortage".
4527 Furthermore, in debugging mode it crashed with a floating point exception.
4528 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4529 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4530 serious problem, Exim now writes to the main and panic logs when this
4531 happens, with details of the failure. It then refuses to accept the
4532 incoming message, giving the message "spool directory problem" or "log
4533 directory problem" with a 421 code for SMTP messages.
4535 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4536 0, 1, and 2 exist, because some OS complain for execs without them (see
4537 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4538 descriptors. However, the code omitted to check that the open succeeded,
4539 causing mysterious errors if for some reason the permissions on /dev/null
4540 got screwed. Now Exim writes a message to the main and panic logs, and
4541 bombs out if it can't open /dev/null.
4543 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4544 interact so that it is all more flexible. It is supposed to remain
4545 backwards compatible. Also added extra_local_interfaces.
4547 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4548 to bomb out with an assertion failure - to the client this appears as a
4549 connection drop. This problem occurs in the part of the code that was taken
4550 from the Samba project. Fortunately, the assertion is in a very simple
4551 function, so I have fixed this by reproducing the function inline in the
4552 one place where it is called, and arranging for authentication to fail
4553 instead of killing the process with assert().
4555 75. The SPA client code was not working when the server requested OEM rather
4556 than Unicode encoding.
4558 76. Added code to make require_files with a specific uid setting more usable in
4559 the case where statting the file as root fails - usually a non-root-mounted
4560 NFS file system. When this happens and the failure is EACCES, Exim now
4561 forks a subprocess and does the per-uid checking as the relevant uid.
4563 77. Added process_log_path.
4565 78. If log_file_path was not explicitly set, a setting of check_log_space or
4566 check_log_inodes was ignored.
4568 79. If a space check for the spool or log partitions fails, the incident is now
4569 logged. Of course, in the latter case the data may get lost...
4571 80. Added the %p formatting code to string_format() so that it can be used to
4572 print addresses in debug_print(). Adjusted all the address printing in the
4573 debugging in store.c to use %p rather than %d.
4575 81. There was a concern that a line of code in smtp_in.c could overflow a
4576 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4577 initially expressed, the concern was not well-founded, because trailing
4578 spaces are removed early. However, if the trailing spaces were followed by
4579 a NULL, they did not get removed, so the overflow was possible. Two fixes
4582 (a) I re-wrote the offending code in a cleaner fashion.
4583 (b) If an incoming SMTP command contains a NULL character, it is rejected
4586 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4587 runs initgroups(), so that if the Exim user is in any additional groups,
4588 they will be used during message reception.
4594 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob