1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.478 2007/02/14 12:22:36 steve Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
63 PH/11 Added control=no_pipelining.
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
72 PH/14 Added log_selector = +pid.
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
76 PH/16 Add ${if forany and ${if forall.
78 PH/17 Added dsn_from option to vary the From: line in DSNs.
80 PH/18 Flush SMTP output before performing a callout, unless control =
81 no_callout_flush is set.
83 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
84 was true (the default) a successful delivery failed to delete the retry
85 item, thus causing premature timeout of the address. The bug is now
88 PH/20 Added hosts_avoid_pipelining to the smtp transport.
90 PH/21 Long custom messages for fakedefer and fakereject are now split up
91 into multiline reponses in the same way that messages for "deny" and
92 other ACL rejections are.
94 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
95 with slight modification.
97 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
98 draft, changing the syntax and factoring some duplicate code".
100 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
101 for deliveries of the second and subsequent messages over the same SMTP
104 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
110 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
111 fixed by 4.65/MH/01 (is this a record?) are fixed:
113 (i) An empty string was always treated as zero by the numeric comparison
114 operators. This behaviour has been restored.
116 (ii) It is documented that the numeric comparison operators always treat
117 their arguments as decimal numbers. This was broken in that numbers
118 starting with 0 were being interpreted as octal.
120 While fixing these problems I realized that there was another issue that
121 hadn't been noticed. Values of message_size_limit (both the global option
122 and the transport option) were treated as octal if they started with 0.
123 The documentation was vague. These values are now always treated as
124 decimal, and I will make that clear in the documentation.
130 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
131 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
134 MH/01 Don't check that the operands of numeric comparison operators are
135 integers when their expansion is in "skipping" mode (fixes bug
136 introduced by 4.64-PH/07).
138 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
139 child addresses, Exim now panics and dies. Previously, because the count
140 is held in a short int, deliveries were likely to be lost. As such a
141 large number of recipients for a single message is ridiculous
142 (performance will be very, very poor), I have chosen to impose a limit
143 rather than extend the field.
149 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
150 leftover -K file (the existence of which was triggered by #402).
151 While we were at it, introduced process PID as part of the -K
152 filename. This should rule out race conditions when creating
155 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
156 processing considerably. Previous code took too long for large mails,
157 triggering a timeout which in turn triggers #401.
159 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
160 in the DK code in transports.c. sendfile() is not really portable,
161 hence the _LINUX specificness.
163 TF/01 In the add_headers option to the mail command in an Exim filter,
164 there was a bug that Exim would claim a syntax error in any
165 header after the first one which had an odd number of characters
168 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
169 callout verification, Exim cached a "reject" for the entire domain. This
170 is correct for most verifications, but it is not correct for a recipient
171 verification with use_sender or use_postmaster set, because in that case
172 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
173 case of MAIL FROM:<> rejection from other early rejections (e.g.
174 rejection of HELO). When verifying a recipient using a non-null MAIL
175 address, the cache is ignored if it shows MAIL FROM:<> rejection.
176 Whatever the result of the callout, the value of the domain cache is
177 left unchanged (for any other kind of callout, getting as far as trying
178 RCPT means that the domain itself is ok).
180 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
183 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
184 manifest itself as EPIPE rather than ECONNECT. When tidying away a
185 session, the daemon ignores ECONNECT errors and logs others; it now
186 ignores EPIPE as well.
188 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
189 (quoted-printable decoding).
191 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
192 later the small subsequent patch to fix an introduced bug.
194 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
196 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
198 PH/08 An error is now given if message_size_limit is specified negative.
200 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
201 to be given (somewhat) arbitrary names.
203 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
206 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
207 miscellaneous code fixes
209 PH/10 Added the log_reject_target ACL modifier to specify where to log
212 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
213 hostname. This is wrong, because it relates to the incoming message (and
214 probably the interface on which it is arriving) and not to the outgoing
215 callout (which could be using a different interface). This has been
216 changed to use the value of the helo_data option from the smtp transport
217 instead - this is what is used when a message is actually being sent. If
218 there is no remote transport (possible with a router that sets up host
219 addresses), $smtp_active_hostname is used.
221 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
222 tweaks were necessary in order to get it to work (see also 21 below):
223 (a) The code assumed that strncpy() returns a negative number on buffer
224 overflow, which isn't the case. Replaced with Exim's string_format()
226 (b) There were several signed/unsigned issues. I just did the minimum
227 hacking in of casts. There is scope for a larger refactoring.
228 (c) The code used strcasecmp() which is not a standard C function.
229 Replaced with Exim's strcmpic() function.
230 (d) The code set only $1; it now sets $auth1 as well.
231 (e) A simple test gave the error "authentication client didn't specify
232 service in request". It would seem that Dovecot has changed its
233 interface. Fortunately there's a specification; I followed it and
234 changed what the client sends and it appears to be working now.
236 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
239 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
240 address A is aliased to B and C, where B exists and C does not. Without
241 -v the output is "A verified" because verification stops after a
242 successful redirection if more than one address is generated. However,
243 with -v the child addresses are also verified. Exim was outputting "A
244 failed to verify" and then showing the successful verification for C,
245 with its parentage. It now outputs "B failed to verify", showing B's
246 parentage before showing the successful verification of C.
248 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
249 look up a TXT record in a specific list after matching in a combined
252 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
253 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
254 they consult the DNS. I had assumed they would set it the way they
255 wanted; and indeed my experiments on Linux seem to show that in some
256 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
257 To be on the safe side, however, I have now made the interface to
258 host_find_byname() similar to host_find_bydns(), with an argument
259 containing the DNS resolver options. The host_find_byname() function now
260 sets these options at its start, just as host_find_bydns() does. The smtp
261 transport options dns_qualify_single and dns_search_parents are passed to
262 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
263 of host_find_byname() use the default settings of RES_DEFNAMES
264 (qualify_single) but not RES_DNSRCH (search_parents).
266 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
267 spool_read_header() do less string testing, by means of a preliminary
268 switch on the second character of optional "-foo" lines. (This is
269 overdue, caused by the large number of possibilities that now exist.
270 Originally there were few.) While I was there, I also converted the
271 str(n)cmp tests so they don't re-test the leading "-" and the first
272 character, in the hope this might squeeze out yet more improvement.
274 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
275 flag allowing group syntax was set by the header_syntax check but not
276 turned off, possible causing trouble later; (2) The flag was not being
277 set at all for the header_verify test, causing "group"-style headers to
278 be rejected. I have now set it in this case, and also caused header_
279 verify to ignore an empty address taken from a group. While doing this, I
280 came across some other cases where the code for allowing group syntax
281 while scanning a header line wasn't quite right (mostly, not resetting
282 the flag correctly in the right place). These bugs could have caused
283 trouble for malformed header lines. I hope it is now all correct.
285 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
286 with the "reply" argument non-NULL. The code, however (which originally
287 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
288 but it didn't always do it. This confused somebody who was copying the
289 code for some other use. I have removed all the tests.
291 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
292 feature that was used to support insecure browsers during the U.S. crypto
293 embargo. It requires special client support, and Exim is probably the
294 only MTA that supported it -- and would never use it because real RSA is
295 always available. This code has been removed, because it had the bad
296 effect of slowing Exim down by computing (never used) parameters for the
297 RSA_EXPORT functionality.
299 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
300 authenticator to fail if there's a tab character in the incoming data
301 (there should never be unless someone is messing about, as it's supposed
302 to be base64-encoded). Also added, on Timo's advice, the "secured" option
303 if the connection is using TLS or if the remote IP is the same as the
304 local IP, and the "valid-client-cert option" if a client certificate has
307 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
308 authenticators. This can be used for authorization after authentication
309 succeeds. (In the case of plaintext, it servers for both authentication
312 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
313 if any retry times were supplied.
315 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
316 connection, where there is no client IP address to check. In this
317 situation, the verify now always succeeds.
319 PH/25 Applied John Jetmore's -Mset patch.
321 PH/26 Added -bem to be like -Mset, but loading a message from a file.
323 PH/27 In a string expansion for a processed (not raw) header when multiple
324 headers of the same name were present, leading whitespace was being
325 removed from all of them, but trailing whitespace was being removed only
326 from the last one. Now trailing whitespace is removed from each header
327 before concatenation. Completely empty headers in a concatenation (as
330 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
331 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
333 PH/29 [Removed. This was a change that I later backed out, and forgot to
334 correct the ChangeLog entry (that I had efficiently created) before
335 committing the later change.]
337 PH/30 Exim was sometimes attempting to deliver messages that had suffered
338 address errors (4xx response to RCPT) over the same connection as other
339 messages routed to the same hosts. Such deliveries are always "forced",
340 so retry times are not inspected. This resulted in far too many retries
341 for the affected addresses. The effect occurred only when there were more
342 hosts than the hosts_max_try setting in the smtp transport when it had
343 the 4xx errors. Those hosts that it had tried were not added to the list
344 of hosts for which the message was waiting, so if all were tried, there
345 was no problem. Two fixes have been applied:
347 (i) If there are any address or message errors in an SMTP delivery, none
348 of the hosts (tried or untried) are now added to the list of hosts
349 for which the message is waiting, so the message should not be a
350 candidate for sending over the same connection that was used for a
351 successful delivery of some other message. This seems entirely
352 reasonable: after all the message is NOT "waiting for some host".
353 This is so "obvious" that I'm not sure why it wasn't done
354 previously. Hope I haven't missed anything, but it can't do any
355 harm, as the worst effect is to miss an optimization.
357 (ii) If, despite (i), such a delivery is accidentally attempted, the
358 routing retry time is respected, so at least it doesn't keep
359 hammering the server.
361 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
362 in ${readsocket because some servers need this prod.
364 PH/32 Added some extra debug output when updating a wait-xxx database.
366 PH/33 The hint "could be header name not terminated by colon", which has been
367 given for certain expansion errors for a long time, was not being given
368 for the ${if def:h_colon_omitted{... case.
370 PH/34 The spec says: "With one important exception, whenever a domain list is
371 being scanned, $domain contains the subject domain." There was at least
372 one case where this was not true.
374 PH/35 The error "getsockname() failed: connection reset by peer" was being
375 written to the panic log as well as the main log, but it isn't really
376 panic-worthy as it just means the connection died rather early on. I have
377 removed the panic log writing for the ECONNRESET error when getsockname()
380 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
381 runs only) independently of the message's sender address. This meant
382 that, if the 4xx error was in fact related to the sender, a different
383 message to the same recipient with a different sender could confuse
384 things. In particualar, this can happen when sending to a greylisting
385 server, but other circumstances could also provoke similar problems.
386 I have changed the default so that the retry time for these errors is now
387 based a combination of the sender and recipient addresses. This change
388 can be overridden by setting address_retry_include_sender=false in the
391 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
392 remote server are returned as part of bounce messages. This was not
393 happening for LMTP over a pipe (the lmtp transport), but now it is the
394 same for both kinds of LMTP.
396 PH/38 Despite being documented as not happening, Exim was rewriting addresses
397 in header lines that were in fact CNAMEs. This is no longer the case.
399 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
400 and queue runs started by the daemon processed all messages. This has
401 been fixed so that -R and -S can now usefully be given with -q<time>.
403 PH/40 Import PCRE release 6.7 (fixes some bugs).
405 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
407 PH/42 Give an error if -q is specified more than once.
409 PH/43 Renamed the variables $interface_address and $interface_port as
410 $received_ip_address and $received_port, to make it clear that these
411 values apply to message reception, and not to the outgoing interface when
412 a message is delivered. (The old names remain recognized, of course.)
414 PH/44 There was no timeout on the connect() call when using a Unix domain
415 socket in the ${readsocket expansion. There now is.
417 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
418 be meaningful with "accept".
420 SC/01 Eximstats V1.43
421 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
423 SC/02 Eximstats V1.44
424 Use a glob alias rather than an array ref in the generated
425 parser. This improves both readability and performance.
427 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
428 Collect SpamAssassin and rejection statistics.
429 Don't display local sender or destination tables unless
430 there is data to show.
431 Added average volumes into the top table text output.
433 SC/04 Eximstats V1.46
434 Collect data on the number of addresses (recipients)
435 as well as the number of messages.
437 SC/05 Eximstats V1.47
438 Added 'Message too big' to the list of mail rejection
439 reasons (thanks to Marco Gaiarin).
441 SC/06 Eximstats V1.48
442 Mainlog lines which have GMT offsets and are too short to
443 have a flag are now skipped.
445 SC/07 Eximstats V1.49 (Alain Williams)
446 Added the -emptyok flag.
448 SC/08 Eximstats V1.50
449 Fixes for obtaining the IP address from reject messages.
451 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
452 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
453 whitesspace changes from 4.64-PH/27
455 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
458 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
459 are found, allow negative numbers in numeric criteria)
461 JJ/06 exipick.20061117.2, added new $message_body_missing variable
463 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
464 to match changes made in 4.64-PH/43
466 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
468 PH/47 Put in an explicit test for a DNS lookup of an address record where the
469 "domain" is actually an IP address, and force a failure. This locks out
470 those revolvers/nameservers that support "A-for-A" lookups, in
471 contravention of the specifications.
473 PH/48 When a host name was looked up from an IP address, and the subsequent
474 forward lookup of the name timed out, the host name was left in
475 $sender_host_name, contrary to the specification.
477 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
478 restricted to single-key lookups, Exim was not diagnosing an error if
479 * or *@ was used with a query-style lookup.
481 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
483 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
484 long ago, but noone interested enough thought of it. Let's just say that
485 the "1.1" means that there are some new functions that weren't there at
486 some point in the past.
488 PH/51 Error processing for expansion failure of helo_data from an smtp
489 transport during callout processing was broken.
491 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
492 tested/used via the -bh/-bhc/-bs options.
494 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
495 bug, fixed in subsequent PCRE releases).
497 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
498 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
500 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
506 SC/01 Use a glob alias rather than an array ref in eximstats generated
507 parser. This improves both readability and performance.
509 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
510 Don't display local sender or destination tables in eximstats unless
511 there is data to show.
512 Added average volumes into the eximstats top table text output.
514 SC/03 Collect data on the number of addresses (recipients) as well
515 as the number of messages in eximstats.
517 TF/01 Correct an error in the documentation for the redirect router. Exim
518 does (usually) call initgroups() when daemonizing.
520 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
521 with consistent privilege compared to when running as a daemon.
523 TF/03 Note in the spec that $authenticated_id is not set for local
524 submissions from trusted users.
526 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
527 Thanks to Dean Brooks <dean@iglou.com> for the patch.
529 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
530 by adding some example configuration directives to the default
531 configuration file. A little bit of work is required to uncomment the
532 directives and define how usernames and passwords are checked, but
533 there is now a framework to start from.
535 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
536 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
537 without this. I don't know how relevant this is to other LDAP libraries.
539 PH/02 Add the verb name to the "unknown ACL verb" error.
541 PH/03 Magnus Holmgren's patch for filter_prepend_home.
543 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
545 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
546 directory not expanded when it should be if an expanded home directory
547 was set for the address (which is overridden by the transport).
549 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
552 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
553 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
554 because it is too late at that time, and has no effect.
556 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
557 security issue with \' (bugzilla #107). I could not use the
558 PQescapeStringConn() function, because it needs a PGconn value as one of
561 PH/08 When testing addresses using -bt, indicate those final addresses that
562 are duplicates that would not cause an additional delivery. At least one
563 person was confused, thinking that -bt output corresponded to deliveries.
564 (Suppressing duplicates isn't a good idea as you lose the information
565 about possibly different redirections that led to the duplicates.)
567 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
568 systems where poll() doesn't work, in particular OS X.
570 PH/10 Added more information to debugging output for retry time not reached.
572 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
573 operations in malware.c.
575 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
578 PH/13 If write_rejectlog was set false when logging was sent to syslog with
579 syslog_duplication set false, log lines that would normally be written
580 both the the main log and to the reject log were not written to syslog at
583 PH/14 In the default configuration, change the use of "message" in ACL warn
584 statements to "add_header".
586 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
587 not followed by a command (e.g. "seen endif").
589 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
590 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
593 PH/17 Added extra conditions to the default value of delay_warning_condition
597 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
598 { match{$h_precedence:}{(?i)bulk|list|junk} } \
599 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
602 The Auto-Submitted: and various List- headers are standardised, whereas I
603 don't think Precedence: ever was.
605 PH/18 Refactored debugging code in route_finduser() to show more information,
606 in particular, the error code if getpwnam() issues one.
608 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
609 This is apparently needed in addition to the PH/07 change above to avoid
610 any possible encoding problems.
612 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
613 but not after initializing Perl.
615 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
616 output them only if debugging. By default they are written stderr,
617 apparently, which is not desirable.
619 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
622 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
625 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
627 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
628 authenticated or an ident call has been made. Suppress the default
629 values for $authenticated_id and $authenticated_sender (but permit -oMai
630 and -oMas) when testing with -bh.
632 PH/24 Re-jigged the order of the tests in the default configuration so that the
633 tests for valid domains and recipients precede the DNS black list and CSA
634 tests, on the grounds that those ones are more expensive.
636 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
637 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
638 This bug exists in every version of Exim that I still have, right back to
641 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
642 However, an attempt to turn on case-sensitivity in a regex key by
643 including (?-i) didn't work because the subject string was already
644 lowercased, and the effects were non-intuitive. It turns out that a
645 one-line patch can be used to allow (?-i) to work as expected.
651 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
652 other effects) broke the use of negated acl sub-conditions.
654 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
657 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
658 "Deny" causes Exim to reject the incoming connection with a 554 error.
659 Unfortunately, if there is a major crisis, such as a disk failure,
660 tcp-wrappers gives "deny", whereas what one would like would be some
661 kind of temporary error. A kludge has been added to help with this.
662 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
663 554 error is used if errno is still zero or contains ENOENT (which occurs
664 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
667 PH/03 Add -lutil to the default FreeBSD LIBS setting.
669 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
670 errors. Otherwise a message that provokes a temporary error (when other
671 messages do not) can cause a whole host to time out.
673 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
674 addresses were routed directly to files or pipes from a redirect router.
675 File deliveries just didn't batch; pipe deliveries might have suffered
678 PH/06 A failure to get a lock for a hints database would erroneously always say
679 "Failed to get write lock", even when it was really a read lock.
681 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
682 of 0600. This has been changed to use the value of the lockfile_mode
683 option (which defaults to 0600).
685 PH/08 Applied small patch from the Sieve maintainer.
687 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
688 folder from quota calculations, a direct delivery into this folder messed
689 up the contents of the maildirsize file. This was because the regex was
690 used only to exclude .Trash (or whatever) when the size of the mailbox
691 was calculated. There was no check that a delivery was happening into an
692 excluded directory. This bug has been fixed by ignoring all quota
693 processing for deliveries into excluded directories.
695 PH/10 Added the maildirfolder_create_regex option to appendfile.
701 PH/01 The code for finding all the local interface addresses on a FreeBSD
702 system running IPv6 was broken. This may well have applied to all BSD
703 systems, as well as to others that have similar system calls. The broken
704 code found IPv4 interfaces correctly, but gave incorrect values for the
705 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
706 that it would not match correctly against @[] and not recognize the IPv6
709 PH/02 The ipliteral router was not recognizing addresses of the form user@
710 [ipv6:....] because it didn't know about the "ipv6:" prefix.
712 PH/03 Added disable_ipv6.
714 PH/04 Changed $reply_address to use the raw form of the headers instead of the
715 decoded form, because it is most often used to construct To: headers
716 lines in autoreplies, and the decoded form may well be syntactically
717 invalid. However, $reply_address has leading white space removed, and all
718 newlines turned into spaces so that the autoreply transport does not
721 PH/05 If group was specified without a user on a router, and no group or user
722 was specified on a transport, the group from the router was ignored.
724 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
725 for visible compile-time settings that can be used to change these
726 numbers, for those that want even more. Backwards compatibility with old
727 spool files has been maintained. However, going back to a previous Exim
728 release will lost any variables that are in spool files.
730 PH/07 Two small changes when running in the test harness: increase delay when
731 passing a TCP/IP connection to a new process, in case the original
732 process has to generate a bounce, and remove special handling of
733 127.0.0.2 (sic), which is no longer necessary.
735 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
736 be the same on different OS.
738 PH/09 Moved a debug statement in filter processing to avoid a race problem when
741 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
742 whether --show-vars was specified or not
744 JJ/02 exipick: Added support for new ACL variable spool format introduced
747 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
748 syntactically invalid From: or Reply-to: line, and a filter used this to
749 generate an autoreply, and therefore failed to obtain an address for the
750 autoreply, Exim could try to deliver to a non-existent relative file
751 name, causing unrelated and misleading errors. What now happens is that
752 it logs this as a hard delivery error, but does not attempt to create a
755 PH/11 The exinext utility has a -C option for testing purposes, but although
756 the given file was scanned by exinext itself; it wasn't being passed on
757 when Exim was called.
759 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
760 an end-of-file indication when reading a command response.
762 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
763 compiled. In many other places in Exim, IPv6 addresses are always
764 recognized, so I have changed this. It also means that IPv4 domain
765 literals of the form [IPV4:n.n.n.n] are now always recognized.
767 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
768 used if the router is not running as root, for example, when verifying at
769 ACL time, or when using -bh. The debugging output from this situation was
770 non-existent - all you got was a failure to exec. I have made two
773 (a) Failures to set uid/gid, the current directory, or a process leader
774 in a subprocess such as that created by queryprogram now generate
775 suitable debugging ouput when -d is set.
777 (b) The queryprogram router detects when it is not running as root,
778 outputs suitable debugging information if -d is set, and then runs
779 the subprocess without attempting to change uid/gid.
781 PH/15 Minor change to Makefile for building test_host (undocumented testing
784 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
785 additional section of a DNS packet that returns MX or SRV records.
786 Instead, it always explicitly searches for A/AAAA records. This avoids
787 major problems that occur when a DNS server includes only records of one
788 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
789 fixed another bug: if SRV records were looked up and the corresponding
790 address records were *not* found in the additional section, the port
791 values from the SRV records were lost.
793 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
794 using the correct key (the original address) when searching the retry
795 rules in order to find which one to use for generating the retry hint.
797 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
798 adding the default one. Similarly, if it contains a Reply-To: header, the
799 errors_reply_to option, if set, is not used.
801 PH/19 When calculating a retry time, Exim used to measure the "time since
802 failure" by looking at the "first failed" field in the retry record. Now
803 it does not use this if it is later than than the arrival time of the
804 message. Instead it uses the arrival time. This makes for better
805 behaviour in cases where some deliveries succeed, thus re-setting the
806 "first failed" field. An example is a quota failure for a huge message
807 when small messages continue to be delivered. Without this change, the
808 "time since failure" will always be short, possible causing more frequent
809 delivery attempts for the huge message than are intended.
810 [Note: This change was subsequently modified - see PH/04 for 4.62.]
812 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
813 $1, $2, $3) because the numerical variables can be reset during some
814 expansion items (e.g. "match"), thereby losing the authentication data.
816 PH/21 Make -bV show the size of off_t variables so that the test suite can
817 decide whether to run tests for quotas > 2G.
819 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
820 mailbox_size, and mailbox_filecount in the appendfile transport. If a
821 filecount value is greater than 2G or if a quota value is greater than 2G
822 on a system where the size of off_t is not greater than 4, a panic error
825 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
826 never match. The debug and -bh output now contains an explicit error
827 message indicating a malformed IPv4 address or mask.
829 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
830 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
833 PH/25 Do not write to syslog when running in the test harness. The only
834 occasion when this arises is a failure to open the main or panic logs
835 (for which there is an explicit test).
837 PH/26 Added the /no_tell option to "control=freeze".
839 PH/27 If a host name lookup failed very early in a connection, for example, if
840 the IP address matched host_lookup and the reverse lookup yielded a name
841 that did not have a forward lookup, an error message of the form "no IP
842 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
843 could be logged. Now it outputs the IP address instead of "NULL".
845 PH/28 An enabling patch from MH: add new function child_open_exim2() which
846 allows the sender and the authenticated sender to be set when
847 submitting a message from within Exim. Since child_open_exim() is
848 documented for local_scan(), the new function should be too.
850 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
851 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
852 results in an empty string is now treated as unset.
854 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
856 PH/31 Added sender_verify_fail logging option.
858 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
859 needed by RFC 822 but not by RFC 2822 was commented out. I have now
860 tidied the source and removed it altogether.
862 PH/33 When a queue run was abandoned because the load average was too high, a
863 log line was always written; now it is written only if the queue_run log
864 selector is set. In addition, the log line for abandonment now contains
865 information about the queue run such as the pid. This is always present
866 in "start" and "stop" lines but was omitted from the "abandon" line.
868 PH/34 Omit spaces between a header name and the colon in the error message that
869 is given when verify = headers_syntax fails (if there are lots of them,
870 the message gets confusing).
872 PH/35 Change the default for dns_check_names_pattern to allow slashes within
873 names, as there are now some PTR records that contain slashes. This check
874 is only to protect against broken name servers that fall over on strange
875 characters, so the fact that it applies to all lookups doesn't matter.
877 PH/36 Now that the new test suite is complete, we can remove some of the
878 special code in Exim that was needed for the old test suite. For example,
879 sorting DNS records because real resolvers return them in an arbitrary
880 order. The new test suite's fake resolver always returns records in the
883 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
884 bounces) except when queue_only is set, to avoid logging races between
885 the different processes.
887 PH/38 Panic-die if .include specifies a non-absolute path.
889 PH/39 A tweak to the "H" retry rule from its user.
891 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
892 a label. They prevented compilation on older perls.
894 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
895 a warning to be raised on newish perls.
897 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
898 on queue. Changes to match documented behaviour of showing count of
899 messages matching specified criteria.
901 PH/40 Changed the default ident timeout from 30s to 5s.
903 PH/41 Added support for the use of login_cap features, on those BSD systems
904 that have them, for controlling the resources used by pipe deliveries.
906 PH/42 The content-scanning code uses fopen() to create files in which to put
907 message data. Previously it was not paying any attention to the mode of
908 the files. Exim runs with umask(0) because the rest of the code creates
909 files with open(), and sets the required mode explicitly. Thus, these
910 files were ending up world-writeable. This was not a big issue, because,
911 being within the spool directory, they were not world-accessible. I have
912 created a function called modefopen, which takes an additional mode
913 argument. It sets umask(777), creates the file, chmods it to the required
914 mode, then resets the umask. All the relevant calls to fopen() in the
915 content scanning code have been changed to use this function.
917 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
918 to 24 hours. This avoids potential overflow problems when processing G
919 and H retry rules. I suspect nobody ever tinkers with this value.
921 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
923 PH/45 When the plaintext authenticator is running as a client, the server's
924 challenges are checked to ensure they are valid base64 strings. By
925 default, the authentication attempt is cancelled if an invalid string is
926 received. Setting client_ignore_invalid_base64 true ignores these errors.
927 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
928 they are received. Thus, the responses can be made to depend on the
929 challenges. If an invalid string is ignored, an empty string is placed in
932 PH/46 Messages that are created by the autoreply transport now contains a
933 References: header, in accordance with RFCs 2822 and 3834.
935 PH/47 Added authenticated_sender_force to the smtp transport.
937 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
939 PH/49 Installed latest patch from the Sieve maintainer.
941 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
942 was also set, the appendfile transport was unnecessarily scanning a
943 directory of message files (e.g. for maildir delivery) to find the count
944 of files (along with the size), even though it did not need this
945 information. It now does the scan only if it needs to find either the
946 size of the count of files.
948 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
950 PH/52 Two bugs concerned with error handling when the smtp transport is
953 (i) Exim was not creating retry information for temporary errors given
954 for individual recipients after the DATA command when the smtp transport
955 was used in LMTP mode. This meant that they could be retried too
956 frequently, and not timed out correctly.
958 (ii) Exim was setting the flag that allows error details to be returned
959 for LMTP errors on RCPT commands, but not for LMTP errors for individual
960 recipients that were returned after the DATA command.
962 PH/53 This is related to PH/52, but is more general: for any failing address,
963 when detailed error information was permitted to be returned to the
964 sender, but the error was temporary, then after the final timeout, only
965 "retry timeout exceeded" was returned. Now it returns the full error as
966 well as "retry timeout exceeded".
968 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
969 do this, and (what is worse) MTAs that accept it.
971 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
972 will now be deprecated.
974 PH/56 New os.c-cygwin from the Cygwin maintainer.
976 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
977 formats (previously only available in exim formats via -bpr, -bpru,
978 and -bpra. Now also available in native and exiqgrep formats)
980 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
981 with very large, slow to parse queues
983 JJ/08 exipick: added ! as generic prefix to negate any criteria format
985 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
987 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
988 responses to authentication challenges, though it was showing the
989 challenges; (ii) I've removed the CR characters from the debug output for
992 PH/58 Allow for the insertion of a newline as well as a space when a string
993 is turned into more than one encoded-word during RFC 2047 encoding. The
994 Sieve code now uses this.
996 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
997 data_4xx, lost_connection, tls_required.
999 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1000 message was being sent as an SMTP response.
1002 PH/61 Add -l and -k options to exicyclog.
1004 PH/62 When verifying, if an address was redirected to one new address, so that
1005 verification continued, and the new address failed or deferred after
1006 having set something in $address_data, the value of $address_data was not
1007 passed back to the ACL. This was different to the case when no
1008 redirection occurred. The value is now passed back in both cases.
1010 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1011 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1012 use login_cap.h, so on its own it isn't the distinguishing feature. The
1013 new name refers directly to the setclassresources() function.
1015 PH/65 Added configuration files for NetBSD3.
1017 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1019 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1020 is preferred over IPv4.
1022 PH/68 The bounce_return_message and bounce_return_body options were not being
1023 honoured for bounces generated during the reception of non-SMTP messages.
1024 In particular, this applied to messages rejected by the ACL. This bug has
1025 been fixed. However, if bounce_return_message is true and bounce_return_
1026 body is false, the headers that are returned for a non-SMTP message
1027 include only those that have been read before the error was detected.
1028 (In the case of an ACL rejection, they have all been read.)
1030 PH/69 The HTML version of the specification is now built in a directory called
1031 spec_html instead of spec.html, because the latter looks like a path with
1032 a MIME-type, and this confuses some software.
1034 PH/70 Catch two compiler warnings in sieve.c.
1036 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1037 function verify_get_ident() calls ip_connect() to connect a socket, but
1038 if the "connect()" function timed out, ip_connect() used to close the
1039 socket. However, verify_get_ident() also closes the socket later, and in
1040 between Exim writes to the log, which may get opened at this point. When
1041 the socket was closed in ip_connect(), the log could get the same file
1042 descriptor number as the socket. This naturally causes chaos. The fix is
1043 not to close the socket in ip_connect(); the socket should be closed by
1044 the function that creates it. There was only one place in the code where
1045 this was missing, in the iplookup router, which I don't think anybody now
1046 uses, but I've fixed it anyway.
1048 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1049 well as to direct DNS lookups. Otherwise the handling of names in host
1050 lists is inconsistent and therefore confusing.
1056 PH/01 Two changes to the default runtime configuration:
1058 (1) Move the checks for relay_from_hosts and authenticated clients from
1059 after to before the (commented out) DNS black list checks.
1061 (2) Add control=submission to the relay_from_hosts and authenticated
1062 clients checks, on the grounds that messages accepted by these
1063 statements are most likely to be submissions.
1065 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1067 (1) Generate an error if the third argument for the ${prvs expansion is
1070 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1073 (3) Reset the variables that are obtained from the first argument of
1074 ${prvscheck and used in the second argument before leaving the code,
1075 because their memory is reclaimed, so using them afterwards may do
1078 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1079 one (it's much easier than Tom thought :-).
1081 (5) Because of (4), we can now allow for the use of $prvscheck_result
1082 inside the third argument.
1084 PH/03 For some reason, the default setting of PATH when running a command from
1085 a pipe transport was just "/usr/bin". I have changed it to
1088 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1089 anything to be listed in the output from -bV.
1091 PH/05 When a filter generated an autoreply, the entire To: header line was
1092 quoted in the delivery log line, like this:
1094 => >A.N.Other <ano@some.domain> <original@ddress> ...
1096 This has been changed so that it extracts the operative address. There
1097 may be more than one such address. If so, they are comma-separated, like
1100 => >ano@some.domain,ona@other.domain <original@ddress> ...
1102 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1103 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1104 not being looked up in the rDNS to get a host name, Exim was showing the
1105 IP address twice in Received: lines, even though the IP addresses were
1106 identical. For example:
1108 Received: from [1.2.3.4] (helo=[1.2.3.4])
1110 However, if the real host name was known, it was omitting the HELO data
1111 if it matched the actual IP address. This has been tidied up so that it
1112 doesn't show the same IP address twice.
1114 PH/07 When both +timestamp and +memory debugging was on, the value given by
1115 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1116 called by the debug printing, thereby overwriting the timestamp buffer.
1117 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1119 PH/08 When the original message was included in an autoreply transport, it
1120 always said "this is a copy of the message, including all the headers",
1121 even if body_only or headers_only was set. It now gives an appropriate
1124 PH/09 Applied a patch from the Sieve maintainer which:
1126 o fixes some comments
1127 o adds the (disabled) notify extension core
1128 o adds some debug output for the result of if/elsif tests
1129 o points to the current vacation draft in the documentation
1130 and documents the missing references header update
1134 o fixes a bug in processing the envelope test (when testing
1135 multiple envelope elements, the last element determinted the
1138 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1139 Electronic Mail") by including:
1141 Auto-submitted: auto-generated
1143 in the messages that it generates (bounce messages and others, such as
1144 warnings). In the case of bounce messages for non-SMTP mesages, there was
1145 also a typo: it was using "Auto_submitted" (underscore instead of
1146 hyphen). Since every message generated by Exim is necessarily in response
1147 to another message, thes have all been changed to:
1149 Auto-Submitted: auto-replied
1151 in accordance with these statements in the RFC:
1153 The auto-replied keyword:
1155 - SHOULD be used on messages sent in direct response to another
1156 message by an automatic process,
1158 - MUST NOT be used on manually-generated messages,
1160 - MAY be used on Delivery Status Notifications (DSNs) and Message
1161 Disposition Notifications (MDNs),
1163 - MUST NOT be used on messages generated by automatic or periodic
1164 processes, except for messages which are automatic responses to
1167 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1168 to the default Received: header definition.
1170 PH/12 Added log selector acl_warn_skipped (default on).
1172 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1173 variables because (a) they are in the wrong storage pool and (b) even if
1174 they were copied, it wouldn't work properly because of the caching.
1176 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1177 checking when decoding. Apparently there are clients that generate
1178 overlong encoded strings. Why am I not surprised?
1180 PH/15 If the first argument of "${if match_address" was not empty, but did not
1181 contain an "@" character, Exim crashed. Now it writes a panic log message
1182 and treats the condition as false.
1184 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1186 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1187 extension "envelope-auth". The code is finished and in agreement with
1188 other implementations, but there is no documentation so far and in fact,
1189 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1190 not changing the active code.
1192 Print executed "if" and "elsif" statements when debugging is used. This
1193 helps a great deal to understand what a filter does.
1195 Document more things not specified clearly in RFC3028. I had all this
1196 sorted out, when out of a sudden new issues came to my mind. Oops."
1198 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1201 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1202 canonical form (as documented). However, after a host name lookup from
1203 the IP address, check_host() was doing a simple string comparison with
1204 addresses acquired from the DNS when checking that the found name did
1205 have the original IP as one of its addresses. Since any found IPv6
1206 addresses are likely to be in abbreviated form, the comparison could
1207 fail. Luckily, there already exists a function for doing the comparison
1208 by converting both addresses to binary, so now that is used instead of
1209 the text comparison.
1211 PH/20 There was another similar case to PH/19, when a complete host name was
1212 given in a host list; looking up its IP address could give an abbreviated
1213 form, whereas the current host's name might or might not be abbreviated.
1214 The same fix has been applied.
1220 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1221 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1224 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1225 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1227 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1229 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1230 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1231 Also, for "Auto-Submitted", treat anything other than "no" as
1232 non-personal, in accordance with RFC 3834. (Previously it treated
1233 anything starting "auto-" as non-personal.)
1235 TF/01 The control=submission/name=... option had a problem with syntax
1236 errors if the name included a slash character. The /name= option
1237 now slurps the rest of the string, so it can include any characters
1238 but it must come last in the list of options (after /sender_retain
1241 PH/05 Some modifications to the interface to the fake nameserver for the new
1249 TK/01 Added the "success_on_redirect" address verification option. See
1250 NewStuff for rationale and an example.
1252 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1254 PH/02 Patch to exigrep to allow it to work on syslog lines.
1256 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1257 fread() to skip over the body file's header line, because in Cygwin the
1258 header line is locked and is inaccessible.
1260 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1261 co-exist for some time) to make it clear that it is the Exim ID that is
1262 referenced, not the Message-ID: header line.
1264 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1265 string_format() function, because snprintf() does not exist on all
1268 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1269 ${stat: expansion item.
1271 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1272 protocol synchronization error", to keep the pedants happy.
1274 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1275 well as for IRIX systems, when gcc is being used. See the host.c source
1278 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1280 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1283 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1284 transport and to the smtp transport in LMTP mode.
1286 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1288 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1289 run a filter in a subprocess. This could lead to confusion in subsequent
1290 lookups in the parent process. There should also be a search_tidyup() at
1291 the end of the subprocess.
1293 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1294 only if the host matched helo_try_verify_hosts, which caused the
1295 verification to occur when the EHLO/HELO command was issued. The ACL just
1296 tested the remembered result. Now, if a previous verification attempt has
1297 not happened, "verify = helo" does it there and then.
1299 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1301 TK/03 Fix log output including CR from clamd.
1303 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1304 exist provoked a memory error which could cause a segfault.
1306 PH/15 Installed PCRE 6.2
1308 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1310 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1311 of the problem. Specifically, suggested +O2 rather than +O1 for the
1314 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1316 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1317 record, the port was not correctly logged when the outgoing_port log
1318 selector was set (it logged the transort's default port).
1320 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1321 fallback_hosts, and "hosts" in the smtp transport.
1323 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1324 host errors such as "Connection refused".
1326 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1327 authentication with radiusclient 0.4.9:
1329 - Error returned from rc_read_config was caught wrongly
1330 - Username/password not passed on to radius server due to wrong length.
1332 The presumption is that some radiusclient API changes for 4.51/PH/17
1333 were not taken care of correctly. The code is still untested by me (my
1334 Linux distribution still has 0.3.2 of radiusclient), but it was
1335 contributed by a Radius user.
1337 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1338 expanding the "port" option of the smtp transport.
1340 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1341 while reading a MIME header. Thanks to Tom Hughes for a patch.
1343 PH/24 Include config.h inside local_scan.h so that configuration settings are
1346 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1347 that in an ACL for RCPT (for example), you can examine exactly what was
1350 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1351 commands, but it was not correctly comparing the address with the actual
1352 client host address. Thus, it would show the EHLO address in Received:
1353 header lines when this was not necessary.
1355 PH/27 Added the % operator to ${eval:}.
1357 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1358 it should be ignoring failures (because with -C, for example, it has lost
1359 privilege). It wasn't ignoring creation failures other than "already
1362 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1365 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1366 because an input line was too long, either on its own, or by virtue of
1367 too many continuations, the temporary file was not being removed, and the
1368 return code was incorrect.
1370 PH/31 Missing "BOOL" in function definition in filtertest.c.
1372 PH/32 Applied Sieve patches from the maintainer.
1374 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1376 PH/33 Added "verify = not_blind".
1378 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1379 Local/Makefile (with some defaults set). These are used in built scripts
1380 such as exicyclog, but they have never been used in the exim_install
1381 script (though there are many overriding facilities there). I have
1382 arranged that the exim_install script now takes note of these two
1385 PH/35 Installed configuration files for Dragonfly.
1387 PH/36 When a locally submitted message by a trusted user did not contain a
1388 From: header, and the sender address was obtained from -f or from an SMTP
1389 MAIL command, and the trusted user did not use -F to supply a sender
1390 name, $originator_name was incorrectly used when constructing a From:
1391 header. Furthermore, $originator_name was used for submission mode
1392 messages from external hosts without From: headers in a similar way,
1393 which is clearly wrong.
1395 PH/37 Added control=suppress_local_fixups.
1397 PH/38 When log_selector = +received_sender was set, and the addition of the
1398 sender made the log line's construction buffer exactly full, or one byte
1399 less than full, an overflow happened when the terminating "\n" was
1402 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1403 when the result of a list match is failure because a DNS lookup failed.
1405 PH/40 RM_COMMAND is now used in the building process.
1407 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1408 the "build-* directories that it finds.
1410 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1411 address in a domain literal was a prefix of an interface address.
1413 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1414 when verifying a sender address, unless rewrite_headers is false.
1416 PH/44 Wrote a long comment about why errors_to addresses are verified as
1417 recipients, not senders.
1419 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1420 the ratelimit ACL was added.
1422 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1424 PH/46 Added extra information about PostgreSQL errors to the error string.
1426 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1427 avoiding the need to install special zones in a real server. This is
1428 backwards compatible; if it can't find the fake resolver, it drops back.
1429 Thus, both old and new test suites can be run.
1431 TF/02 Added util/ratelimit.pl
1433 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1434 clock is set back in time.
1436 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1437 Candler <B.Candler@pobox.com>.
1439 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1440 OK for addresses that are the result of redirections.
1442 PH/48 A number of further additions for the benefit of the new test suite,
1443 including a fake gethostbyname() that interfaces to the fake DNS resolver
1446 TF/06 The fix for widen_domains has also been applied to qualify_single and
1447 search_parents which are the other dnslookup options that can cause
1450 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1453 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1455 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1456 service" if it encountered an MX record with an empty target hostname.
1457 The message is now "an MX or SRV record indicated no SMTP service".
1459 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1460 if the DNS of the sending site is misconfigured. This is quite a
1461 common situation. This change restores the behaviour of treating a
1462 helo verification defer as a failure.
1464 PH/51 If self=fail was set on a router, the bounce message did not include the
1465 actual error message.
1471 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1473 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1474 command itself ended in error, the underlying message about the transport
1475 filter timeout was being overwritten with the pipe command error. Now the
1476 underlying error message should be appended to the second error message.
1478 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1480 PH/02 Reduce the amount of output that "make" produces by default. Full output
1481 can still be requested.
1483 PH/03 The warning log line about a condition test deferring for a "warn" verb
1484 was being output only once per connection, rather than after each
1485 occurrence (because it was using the same function as for successful
1486 "warn" verbs). This seems wrong, so I have changed it.
1488 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1489 it should not have, which might have caused a crash in the right
1490 circumstances, but probably never did.
1492 PH/04 Installed a modified version of Tony Finch's patch to make submission
1493 mode fix the return path as well as the Sender: header line, and to
1494 add a /name= option so that you can make the user's friendly name appear
1497 TF/03 Added the control = fakedefer ACL modifier.
1499 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1500 Mark Lowes for thorough testing.
1502 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1504 TK/03 Merged latest SRS patch from Miles Wilton.
1506 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1507 with the definition in sysexits.h (which is #included earlier).
1508 Fortunately, Exim does not actually use EX_OK. The code used to try to
1509 preserve the sysexits.h value, by assumimg that macro definitions were
1510 scanned for macro replacements. I have been disabused of this notion,
1511 so now the code just undefines EX_OK before #including unistd.h.
1513 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1514 in the smtp transport. When a block could not be written in a single
1515 write() function, the timeout was being re-applied to each part-write.
1516 This seems wrong - if the receiver was accepting one byte at a time it
1517 would take for ever. The timeout is now adjusted when this happens. It
1518 doesn't have to be particularly precise.
1520 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1521 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1523 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1524 without a domain if the check to <postmaster@domain> fails.
1526 SC/01 Eximstats: added -xls and the ability to specify output files
1527 (patch written by Frank Heydlauf).
1529 SC/02 Eximstats: use FileHandles for outputing results.
1531 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1533 SC/04 Eximstats: fixed display of large numbers with -nvr option
1535 SC/05 Eximstats: fixed merging of reports with empty tables.
1537 SC/06 Eximstats: added the -include_original_destination flag
1539 SC/07 Eximstats: removed tabs and trailing whitespace.
1541 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1543 TK/06 MBOX spool code: Add real "From " MBOX separator line
1544 so the .eml file is really in mbox format (even though
1545 most programs do not really care). Patch from Alex Miller.
1547 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1548 The latter is generated from $received_to and is only set if the
1549 message has one envelope recipient. SA can use these headers,
1550 obviously out-of-the-box. Patch from Alex Miller.
1552 PH/08 The ${def test on a variable was returning false if the variable's
1553 value was "0", contrary to what the specification has always said!
1554 The result should be true unless the variable is empty.
1556 PH/09 The syntax error of a character other than { following "${if
1557 def:variable_name" (after optional whitespace) was not being diagnosed.
1558 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1559 accidental colon was present, for example, could give incorrect results.
1561 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1562 result is used (not including appendfile, where other changes are about
1565 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1566 This involved changing a lot of size variables from int to off_t. It
1567 should work with maildirs and everything.
1569 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1570 spamd dying while we are connected to it.
1572 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1575 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1576 function for BDB 4.3.
1578 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1580 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1581 the sources for PCRE have been re-arranged and more files are now
1584 PH/15 The code I had for printing potentially long long variables in PH/11
1585 above was not the best (it lost precision). The length of off_t variables
1586 is now inspected at build time, and an appropriate printing format (%ld
1587 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1588 to be "long long int" or "long int". This is needed for the internal
1589 formatting function string_vformat().
1591 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1592 the configuration file to be ":syslog", then the script "guesses" where
1593 the logs files are, rather than using the compiled in default. In our
1594 case the guess is not the same as the compiled default, so the script
1595 suddenly stopped working when I started to use syslog. The patch checks
1596 to see if log_file_path is "". If so, it attempts to read it from exim
1597 with no configuration file to get the compiled in version, before it
1598 falls back to the previous guessing code."
1600 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1601 implementing BATV in an Exim configuration. See NewStuff for the gory
1604 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1605 Makefile that are specific to HP-UX.
1607 PH/18 If the "use_postmaster" option was set for a recipient callout together
1608 with the "random" option, the postmaster address was used as the MAIL
1609 FROM address for the random test, but not for the subsequent recipient
1610 test. It is now used for both.
1612 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1613 patch removes a few documentation additions to RFC 3028, because the
1614 latest draft now contains them. It adds the new en;ascii-case comparator
1615 and a new error check for 8bit text in MIME parts. Comparator and
1616 require names are now matched exactly. I enabled the subaddress
1617 extension, but it is not well tested yet (read: it works for me)."
1619 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1620 rework some of the code of TK/09 above to avoid the hardwired use of
1621 "%lld" and "long long". Replaced the call to snprintf() with a call to
1624 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1625 records point to non-existent hosts", "retry timeout exceeded", and
1626 "retry time not reached for any host after a long failure period".
1628 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1629 experimental DomainKeys support:
1631 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1632 (2) On an error such as an illegally used "control", the wrong name for
1633 the control was given.
1635 These problems did NOT occur unless DomainKeys support was compiled.
1637 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1639 PH/24 Added ${if match_ip condition.
1641 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1642 databases so that it will be absolutely obvious if a crash occurs in the
1643 DB library. This is a regular occurrence (often caused by mis-matched
1646 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1647 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1648 picked up on a user's system that detects such things. There doesn't seem
1649 to be a gcc warning option for this - only an attribute that has to be
1650 put on the function's prototype. It seems that in Fedora Core 4 they have
1651 set this on a number of new functions. No doubt there will be more in due
1654 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1655 specify a transport. However, if an address that was verified by such a
1656 router was the subject of a callout, Exim crashed because it tried to
1657 read the rcpt_include_affixes from the non-existent transport. Now it
1658 just assumes that the setting of that option is false. This bug was
1659 introduced by 4.51/PH/31.
1661 PH/28 Changed -d+all to exclude +memory, because that information is very
1662 rarely of interest, but it makes the output a lot bigger. People tend to
1663 do -d+all out of habit.
1665 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1666 code in os-type was giving problems when libc.so lives in lib64, like on
1669 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1670 aren't the modern standard, and it seems that some systems' include files
1671 don't always have them. Exim was already checking for some of the newer
1672 ones like T_AAAA, and defining it itself. I've added checks for all the
1673 record types that Exim uses.
1675 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1676 not automatically generating a new one, as it is supposed to. This
1677 prevented TLS from working. If the file did exist, but contained invalid
1678 data, a new version was generated, as expected. It was only the case of a
1679 non-existent file that was broken.
1681 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1682 with a change in libdomainkeys > 0.64.
1684 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1685 from DNS. If the selector record carries the flag, it now has
1686 precedence over the domain-wide flag.
1688 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1690 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1691 the use of a port name that isn't defined in /etc/services) occurred, the
1692 message was deferred as in a normal delivery, and thus remained on the
1693 spool, instead of being failed because of the mua_wrapper setting. This
1694 is now fixed, and I tidied up some of the mua_wrapper messages at the
1697 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1698 the messages in a hash of arrays rather than using individual hashes.
1699 This is a bit cleaner and results in dramatic memory savings, albeit
1700 at a slight CPU cost.
1702 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1703 as requested by Marc Sherman.
1705 SC/10 Eximstats: added histograms for user specified patterns as requested
1708 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1710 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1711 fopen() in the content-scanning modules that did not already have it.
1717 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1718 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1720 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1722 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1724 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1725 to test Sieve filters that use "vacation".
1727 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1728 that changes the way the GnuTLS parameters are stored in the cache file.
1729 The new format can be generated externally. For backward compatibility,
1730 if the data in the cache doesn't make sense, Exim assumes it has read an
1731 old-format file, and it generates new data and writes a new file. This
1732 means that you can't go back to an older release without removing the
1735 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1736 work if there are any delivery delays because "one_time" forces the
1737 parent to be marked "delivered", so its unseen clone is never tried
1738 again. For this reason, Exim now forbids the simultaneous setting of
1741 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1742 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1743 entry if you want to know the details. The fix, however, neglected to
1744 consider the case when local delivery batching is involved. The test for
1745 "previously delivered" was not happening when checking to see if an
1746 address could be batched with a previous (undelivered) one; under
1747 certain circumstances this could lead to multiple deliveries to the same
1750 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1751 in its include files, and this causes problems building Exim.
1753 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1754 header_syntax) but Exim was just ignoring anything given after a slash.
1755 In particular, this caused confusion with an attempt to use "verify =
1756 reverse_host_lookup/defer_ok". An error is now given when options are
1757 supplied for verify items that do not have them. (Maybe reverse_host_
1758 lookup should have a defer_ok option, but that's a different point.)
1760 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1761 defined by RFC 821) to 2048, because there were problems with some AUTH
1762 commands, and RFC 1869 says the size should be increased for extended
1763 SMTP commands that take arguments.
1765 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1768 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1769 "unknown" error; now it says that the functionality isn't in the binary.
1771 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1772 an address' error message when a string expansion fails (syntax or
1773 whatever). Otherwise the password may appear in the log. Following change
1774 PH/42 below, there is no longer a chance of it appearing in a bounce
1777 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1779 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1780 of its addresses was ever tried. (Bugzilla bug #2.)
1782 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1783 the result incorrectly in the debug output. (It correctly added a newline
1784 to what was transported.)
1786 TF/01 Added $received_time.
1788 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1789 commented out examples of how to interface to a virus scanner and to
1790 SpamAssassin. Also added commented examples of av_scanner and
1791 spamd_address settings.
1793 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1794 and controls are allowed in which ACLs. There were a couple of minor
1795 errors. Some of the entries in the conditions table (which is a table of
1796 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1797 negation of where the condition IS allowed.
1799 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1801 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1802 header file does not have a version number, so I've had to invent a new
1803 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1804 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1805 radiusclient), but it was contributed by a Radius user.
1807 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1808 files or directories, for OpenSSL.
1810 PH/19 When an Exim process that is running as root has to create an Exim log
1811 file, it does so in a subprocess that runs as exim:exim so as to get the
1812 ownership right at creation (otherwise, other Exim processes might see
1813 the file with the wrong ownership). There was no test for failure of this
1814 fork() call, which would lead to the process getting stuck as it waited
1815 for a non-existent subprocess. Forks do occasionally fail when resources
1816 run out. I reviewed all the other calls to fork(); they all seem to check
1819 PH/20 When checking for unexpected SMTP input at connect time (before writing
1820 the banner), Exim was not dealing correctly with a non-positive return
1821 from the read() function. If the client had disconnected by this time,
1822 the result was a log entry for a synchronization error with an empty
1823 string after "input=" when read() returned zero. If read() returned -1
1824 (an event I could not check), uninitialized data bytes were printed.
1825 There were reports of junk text (parts of files, etc) appearing after
1828 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1830 PH/22 Added support for macro redefinition, and (re)definition in between
1831 driver and ACL definitions.
1833 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1834 forgetting to use the resulting value; it was using the unexpanded value.
1836 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1837 hadn't been configured. The fix is from Juergen Kreileder, who
1838 understands it better than I do:
1840 "Here's what I see happening with three configured cyrus_sasl
1841 authenticators configured (plain, login, cram-md5):
1843 On startup auth_cyrus_sasl_init() gets called for each of these.
1844 This means three calls to sasl_listmech() without a specified mech_list.
1845 => SASL tests which mechs of all available mechs actually work
1846 => three warnings about OTP not working
1847 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1849 With the patch, sasl_listmech() also gets called three times. But now
1850 SASL's mech_list option is set to the server_mech specified in the the
1851 authenticator. Or in other words, the answer from sasl_listmech()
1852 gets limited to just the mech you're testing for (which is different
1854 => the return list contains just 'plain' or 'login', 'cram-md5' or
1855 nothing depending on the value of ob->server_mech.
1857 I've just tested the patch: Authentication still works fine,
1858 unavailable mechs specified in the exim configuration are still
1859 caught, and the auth.log warnings about OTP are gone."
1861 PH/25 When debugging is enabled, the contents of the command line are added
1862 to the debugging output, even when log_selector=+arguments is not
1865 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1866 answer is "GNU", and only if the return is "GNU/something" is the answer
1869 PH/27 $acl_verify_message is now set immediately after the failure of a
1870 verification in an ACL, and so is available in subsequent modifiers. In
1871 particular, the message can be preserved by coding like this:
1873 warn !verify = sender
1874 set acl_m0 = $acl_verify_message
1876 Previously, $acl_verify_message was set only while expanding "message"
1877 and "log_message" when a very denied access.
1879 PH/28 Modified OS/os.c-Linux with
1881 -#ifndef OS_LOAD_AVERAGE
1882 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1884 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1885 nomenclature these days.)
1887 PH/29 Installed patch from the Sieve maintainer that adds the options
1888 sieve_useraddress and sieve_subaddress to the redirect router.
1890 PH/30 In these circumstances:
1891 . Two addresses routed to the same list of hosts;
1892 . First host does not offer TLS;
1893 . First host accepts first address;
1894 . First host gives temporary error to second address;
1895 . Second host offers TLS and a TLS session is established;
1896 . Second host accepts second address.
1897 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1898 and peerdn, if requested) that were in fact used only for the second
1901 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1902 attention to any local part prefix or suffix that was matched by the
1903 router that accepted the address. It now behaves in the same way as it
1904 does for delivery: the affixes are removed from the local part unless
1905 rcpt_include_affixes is set on the transport.
1907 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1908 timeout during the DATA phase of an incoming message.
1910 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1911 applied a patch sanctioned by the Sieve maintainer.
1913 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1914 the uid or gid is negative. A case of a negative gid caused this to be
1915 noticed. The fix allows for either to be negative.
1917 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1918 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1921 PH/36 The change PH/12 above was broken. Fixed it.
1923 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1924 the grounds that routing the same address twice would always produce the
1925 same answer. This might have been true once, but it is certainly no
1926 longer true now. Routing a child address may depend on the previous
1927 routing that produced that child. Some complicated redirection strategies
1928 went wrong when messages had multiple recipients, and made Exim's
1929 behaviour dependent on the order in which the addresses were given.
1931 I have moved the duplicate checking until after the routing is complete.
1932 Exim scans the addresses that are assigned to local and remote
1933 transports, and removes any duplicates. This means that more work will be
1934 done, as duplicates will always all be routed, but duplicates are
1935 presumably rare, so I don't expect this is of any significance.
1937 For deliveries to pipes, files, and autoreplies, the duplicate checking
1938 still happens during the routing process, since they are not going to be
1941 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1942 It corrects a timeout issue with spamd. This is Ian's comment: "The
1943 background is that sometimes spamd either never reads data from a
1944 connection it has accepted, or it never writes response data. The exiscan
1945 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1946 blindly assumes that writes won't block so it may never time out."
1948 PH/39 Allow G after quota size as well as K and M.
1950 PH/40 The value set for $authenticated_id in an authenticator may not contain
1951 binary zeroes or newlines because the value is written to log lines and
1952 to spool files. There was no check on this. Now the value is run through
1953 the string_printing() function so that such characters are converted to
1954 printable escape sequences.
1956 PH/41 $message_linecount is a new variable that contains the total number of
1957 lines in the message. Compare $body_linecount, which is the count for the
1960 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1961 bounce and delay warning messages, except in certain special cases, which
1964 (a) An SMTP error message from a remote host;
1965 (b) A message specified in a :fail: redirection;
1966 (c) A message specified in a "fail" command in a system filter;
1967 (d) A message specified in a FAIL return from the queryprogram router;
1968 (e) A message specified by the cannot_route_message router option.
1970 In these cases only, Exim does include the error details in bounce and
1971 warning messages. There are also a few cases where bland messages such
1972 as "unrouteable address" or "local delivery error" are given.
1974 PH/43 $value is now also set for the "else" part of a ${run expansion.
1976 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1977 being worked on, but at least Exim now implements the latest version to
1980 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1981 process to complete was treated as a delivery failure, a timeout while
1982 writing the message to the pipe was logged, but erroneously treated as a
1983 successful delivery. Such timeouts include transport filter timeouts. For
1984 consistency with the overall process timeout, these timeouts are now
1985 treated as errors, giving rise to delivery failures by default. However,
1986 there is now a new Boolean option for the pipe transport called
1987 timeout_defer, which, if set TRUE, converts the failures into defers for
1988 both kinds of timeout. A transport filter timeout is now identified in
1991 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1992 systems where "make" and "gmake" are different, calling "gmake" at top
1993 level broke things. I've arranged for the value of $(MAKE) to be passed
1994 from the Makefile to this script so that it can call the same version of
1998 A note about Exim versions 4.44 and 4.50
1999 ----------------------------------------
2001 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2002 changes of various kinds. As a consequence, a big documentation update was
2003 needed. This delayed the release for rather longer than seemed good, especially
2004 in the light of a couple of (minor) security issues. Therefore, the changes
2005 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2006 release. So 4.44 and 4.50 are in effect two different branches that both start
2009 I have left the 4.50 change log unchanged; it contains all the changes since
2010 4.43. The change log for 4.44 is below; many of its items are identical to
2011 those for 4.50. This seems to be the most sensible way to preserve the
2012 historical information.
2018 1. Minor wording change to the doc/README.SIEVE file.
2020 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2021 computation of the current number of files was incorrect.
2023 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2024 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2027 4. Give more explanation in the error message when the command for a transport
2028 filter fails to execute.
2030 5. There are several places where Exim runs a non-Exim command in a
2031 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2032 was being done only for the command run by the queryprogram router. It is
2033 now done for all such subprocesses. The other cases are: ${run, transport
2034 filters, and the commands run by the lmtp and pipe transports.
2036 6. Added CONFIGURE_GROUP build-time option.
2038 7. Some older OS have a limit of 256 on the maximum number of file
2039 descriptors. Exim was using setrlimit() to set 1000 as a large value
2040 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2041 systems. I've change it so that if it can't get 1000, it tries for 256.
2043 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2044 was an oversight, and furthermore, ever since the addition of extra
2045 controls (e.g. 4.43/32), the checks on when to allow different forms of
2046 "control" were broken. There should now be diagnostics for all cases when a
2047 control that does not make sense is encountered.
2049 9. Added the /retain_sender option to "control=submission".
2051 10. $recipients is now available in the predata ACL (oversight).
2053 11. Tidy the search cache before the fork to do a delivery from a message
2054 received from the command line. Otherwise the child will trigger a lookup
2055 failure and thereby defer the delivery if it tries to use (for example) a
2056 cached ldap connection that the parent has called unbind on.
2058 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2059 of $address_data from the recipient verification was clobbered by the
2060 sender verification.
2062 13. The value of address_data from a sender verification is now available in
2063 $sender_address_data in subsequent conditions in the ACL statement.
2065 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2067 15. Added a new option "connect=<time>" to callout options, to set a different
2070 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2071 was its contents. (It was OK if the option was not defined at all.)
2073 17. A "Completed" log line is now written for messages that are removed from
2074 the spool by the -Mrm option.
2076 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2077 information about exactly what failed.
2079 19. Added -dd to debug only the daemon process.
2081 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2082 handles timeouts, both on the server side and network timeouts. Renamed the
2083 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2085 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2086 It is now set to "smtps".
2088 22. $host_address is now set to the target address during the checking of
2089 ignore_target_hosts.
2091 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2092 being passed; this would have caused $sender_host_name to have been used if
2093 matching the list had actually called for a host name (not very likely,
2094 since this list is usually IP addresses). A host name is now passed as
2097 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2098 code that specifies a non-restarting handler (typically sigaction() in
2099 modern systems) in an attempt to fix a rare and obscure crash bug.
2101 25. Narrowed the window for a race in the daemon that could cause it to ignore
2102 SIGCHLD signals. This is not a major problem, because they are used only to
2103 wake it up if nothing else does.
2105 26. A malformed maildirsize file could cause Exim to calculate negative values
2106 for the mailbox size or file count. Odd effects could occur as a result.
2107 The maildirsize information is now recalculated if the size or filecount
2110 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2111 support for a long time. Removed HAVE_SYS_VFS_H.
2113 28. Installed the latest version of exipick from John Jetmore.
2115 29. In an address list, if the pattern was not a regular expression, an empty
2116 subject address (from a bounce message) matched only if the pattern was an
2117 empty string. Non-empty patterns were not even tested. This was the wrong
2118 because it is perfectly reasonable to use an empty address as part of a
2119 database query. An empty address is now tested by patterns that are
2120 lookups. However, all the other forms of pattern expect the subject to
2121 contain a local part and a domain, and therefore, for them, an empty
2122 address still always fails if the pattern is not itself empty.
2124 30. Exim went into a mad DNS loop when attempting to do a callout where the
2125 host was specified on an smtp transport, and looking it up yielded more
2126 than one IP address.
2128 31. Re-factored the code for checking spool and log partition space into a
2129 function that finds that data and another that does the check. The former
2130 is then used to implement four new variables: $spool_space, $log_space,
2131 $spool_inodes, and $log_inodes.
2133 32. The RFC2047 encoding function was originally intended for short strings
2134 such as real names; it was not keeping to the 75-character limit for
2135 encoded words that the RFC imposes. It now respects the limit, and
2136 generates multiple encoded words if necessary. To be on the safe side, I
2137 have increased the buffer size for the ${rfc2047: expansion operator from
2140 33. It is now permitted to omit both strings after an "if" condition; if the
2141 condition is true, the result is "true". As before, when the second string
2142 is omitted, a false condition yields an empty string. This makes it less
2143 cumbersome to write custom ACL and router conditions.
2145 34. Failure to deliver a bounce message always caused it to be frozen, even if
2146 there was an errors_to setting on the router. The errors_to setting is now
2149 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2150 canonical form (fully expanded) before being placed in
2151 $sender_host_address.
2153 36. The table in the code that translates DNS record types into text (T_A to
2154 "A" for instance) was missing entries for NS and CNAME. It is just possible
2155 that this could have caused confusion if both these types were looked up
2156 for the same domain, because the text type is used as part of Exim's
2157 per-process caching. But the chance of anyone hitting this buglet seems
2160 37. The dnsdb lookup has been extended in a number of ways.
2162 (1) There is a new type, "zns", which walks up the domain tree until it
2163 finds some nameserver records. It should be used with care.
2165 (2) There is a new type, "mxh", which is like "mx" except that it returns
2166 just the host names, not the priorities.
2168 (3) It is now possible to give a list of domains (or IP addresses) to be
2169 looked up. The behaviour when one of the lookups defers can be
2170 controlled by a keyword.
2172 (4) It is now possible to specify the separator character for use when
2173 multiple records are returned.
2175 38. The dnslists ACL condition has been extended: it is now possible to supply
2176 a list of IP addresses and/or domains to be looked up in a particular DNS
2179 39. Added log_selector=+queue_time_overall.
2181 40. When running the queue in the test harness, wait just a tad after forking a
2182 delivery process, to get repeatability of debugging output.
2184 41. Include certificate and key file names in error message when GnuTLS fails
2185 to set them up, because the GnuTLS error message doesn't include the name
2186 of the failing file when there is a problem reading it.
2188 42. Allow both -bf and -bF in the same test run.
2190 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2192 44. The "Exiscan patch" is now merged into the mainline Exim source.
2194 45. Sometimes the final signoff response after QUIT could fail to get
2195 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2196 < 0 before doing a fflush(). This bug looks as though it goes back to the
2197 introduction of TLS in release 3.20, but "sometimes" must have been rare
2198 because the tests only now provoked it.
2200 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2201 (this can affect the format of dates).
2203 47. exim_tidydb, when checking for the continued existence of a message for
2204 which it has found a message-specific retry record, was not finding
2205 messages that were in split spool directories. Consequently, it was
2206 deleting retry records that should have stayed in existence.
2208 48. Steve fixed some bugs in eximstats.
2210 49. The SPA authentication driver was not abandoning authentication and moving
2211 on to the next authenticator when an expansion was forced to fail,
2212 contradicting the general specification for all authenticators. Instead it
2213 was generating a temporary error. It now behaves as specified.
2215 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2216 (the order specifies the preference for clients). The order is now AES256,
2217 AES128, 3DES, ARCFOUR128.
2219 51. Small patch to Sieve code - explicitly set From: when generating an
2222 52. Exim crashed if a remote delivery caused a very long error message to be
2223 recorded - for instance if somebody sent an entire SpamAssassin report back
2224 as a large number of 550 error lines. This bug was coincidentally fixed by
2225 increasing the size of one of Exim's internal buffers (big_buffer) that
2226 happened as part of the Exiscan merge. However, to be on the safe side, I
2227 have made the code more robust (and fixed the comments that describe what
2230 53. Now that there can be additional text after "Completed" in log lines (if
2231 the queue_time_overall log selector is set), a one-byte patch to exigrep
2232 was needed to allow it to recognize "Completed" as not the last thing in
2235 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2236 patch that reportedly fixes this has been added. I am not expert enough to
2237 create a test for it. This is what the patch creator wrote:
2239 "I found a little strange behaviour of ldap code when working with
2240 Windows 2003 AD Domain, where users was placed in more than one
2241 Organization Units. When I tried to give exim partial DN, the exit code
2242 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2243 But simultaneously result of request was absolutely normal ldap result,
2244 so I produce this patch..."
2246 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2247 REFERENCE, so I have modified the code to exclude the patch when that macro
2250 55. Some experimental protocols are using DNS PTR records for new purposes. The
2251 keys for these records are domain names, not reversed IP addresses. The
2252 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2253 leaves it alone. Component reversal etc. now happens only for IP addresses.
2256 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2258 57. Double the size of the debug message buffer (to 2048) so that more of very
2259 long debug lines gets shown.
2261 58. The exicyclog utility now does better if the number of log files to keep
2262 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2264 59. Two changes related to the smtp_active_hostname option:
2266 (1) $smtp_active_hostname is now available as a variable.
2267 (2) The default for smtp_banner uses $smtp_active_hostname instead
2268 of $primary_hostname.
2270 60. The host_aton() function is supposed to be passed a string that is known
2271 to be a valid IP address. However, in the case of IPv6 addresses, it was
2272 not checking this. This is a hostage to fortune. Exim now panics and dies
2273 if the condition is not met. A case was found where this could be provoked
2274 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2275 components; fortuitously, this particular loophole had already been fixed
2276 by change 4.50/55 above.
2278 If there are any other similar loopholes, the new check in host_aton()
2279 itself should stop them being exploited. The report I received stated that
2280 data on the command line could provoke the exploit when Exim was running as
2281 exim, but did not say which command line option was involved. All I could
2282 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2283 running as the user.
2286 61. There was a buffer overflow vulnerability in the SPA authentication code
2287 (which came originally from the Samba project). I have added a test to the
2288 spa_base64_to_bits() function which I hope fixes it.
2291 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2292 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2294 63. The daemon start-up calls getloadavg() while still root for those OS that
2295 need the first call to be done as root, but it missed one case: when
2296 deliver_queue_load_max is set with deliver_drop_privilege. This is
2297 necessary for the benefit of the queue runner, because there is no re-exec
2298 when deliver_drop_privilege is set.
2300 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2301 This has been fixed.
2303 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2304 was in use, was not putting the data itself into the right store pool;
2305 consequently, it could be overwritten for a subsequent message in the same
2306 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2309 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2311 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2312 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2313 the function were treating the return as a boolean value, which happened to
2314 work because 0=false and not-0=true, but is not correct code.
2316 68. The host_aton() function was not handling scoped IPv6 addresses (those
2317 with, for example, "%eth0" on the end) correctly.
2319 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2320 negated items (that is, ~something) in unsigned ints. Some compilers
2321 apparently mutter when there is no cast.
2323 70. If an address verification called from an ACL failed, and did not produce a
2324 user-specific message (i.e. there was only a "system" message), nothing was
2325 put in $acl_verify_message. In this situation, it now puts the system
2328 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2329 session; change 4.31/43 added the unwanted input to the log line - except
2330 that it did not do this in the start of session case. It now does.
2332 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2333 This is wrong and can cause the other end to generate a synchronization
2334 error if it is another Exim or anything else that does the synchronization
2335 check. A QUIT command is no longer sent after a timeout.
2337 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2338 during host lookups.
2340 74. The defer_ok option of callout verification was not working if it was used
2341 when verifying addresses in header lines, that is, for this case:
2343 verify = header_sender/callout=defer_ok
2345 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2346 those file descriptors could be used for SMTP connections. If anything
2347 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2348 could be sent to the SMTP client, causing chaos. The daemon now opens
2349 stdin, stdout, and stderr to /dev/null when it puts itself into the
2352 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2353 log by default. The user can override this with suitable Perl magic.
2355 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2356 the log message when discard triggers, was not working for the DATA ACL or
2357 for the non-SMTP ACL.
2359 78. Error message wording change in sieve.c.
2361 79. If smtp_accept_max_per_host was set, the number of connections could be
2362 restricted to fewer than expected, because the daemon was trying to set up
2363 a new connection before checking whether the processes handling previous
2364 connections had finished. The check for completed processes is now done
2365 earlier. On busy systems, this bug wouldn't be noticed because something
2366 else would have woken the daemon, and it would have reaped the completed
2369 80. If a message was submitted locally by a user whose login name contained one
2370 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2371 It caused a spool format error. I have fixed the spool reading code. A
2372 related problem was that the "from" clause in the Received: line became
2373 illegal because of the space(s). It is now covered by ${quote_local_part.
2375 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2378 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2380 83. Patch from Sieve maintainer to fix unterminated string problem in
2381 "vacation" handling.
2383 84. Some minor changes to the Linux configuration files to help with other
2384 OS variants using glibc.
2386 85. One more patch for Sieve to update vacation handling to latest spec.
2389 ----------------------------------------------------
2390 See the note above about the 4.44 and 4.50 releases.
2391 ----------------------------------------------------
2397 1. Change 4.43/35 introduced a bug that caused file counts to be
2398 incorrectly computed when quota_filecount was set in an appendfile
2401 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2402 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2405 3. Give more explanation in the error message when the command for a transport
2406 filter fails to execute.
2408 4. There are several places where Exim runs a non-Exim command in a
2409 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2410 was being done only for the command run by the queryprogram router. It is
2411 now done for all such subprocesses. The other cases are: ${run, transport
2412 filters, and the commands run by the lmtp and pipe transports.
2414 5. Some older OS have a limit of 256 on the maximum number of file
2415 descriptors. Exim was using setrlimit() to set 1000 as a large value
2416 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2417 systems. I've change it so that if it can't get 1000, it tries for 256.
2419 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2420 was an oversight, and furthermore, ever since the addition of extra
2421 controls (e.g. 4.43/32), the checks on when to allow different forms of
2422 "control" were broken. There should now be diagnostics for all cases when a
2423 control that does not make sense is encountered.
2425 7. $recipients is now available in the predata ACL (oversight).
2427 8. Tidy the search cache before the fork to do a delivery from a message
2428 received from the command line. Otherwise the child will trigger a lookup
2429 failure and thereby defer the delivery if it tries to use (for example) a
2430 cached ldap connection that the parent has called unbind on.
2432 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2433 of $address_data from the recipient verification was clobbered by the
2434 sender verification.
2436 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2437 was its contents. (It was OK if the option was not defined at all.)
2439 11. A "Completed" log line is now written for messages that are removed from
2440 the spool by the -Mrm option.
2442 12. $host_address is now set to the target address during the checking of
2443 ignore_target_hosts.
2445 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2446 being passed; this would have caused $sender_host_name to have been used if
2447 matching the list had actually called for a host name (not very likely,
2448 since this list is usually IP addresses). A host name is now passed as
2451 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2452 code that specifies a non-restarting handler (typically sigaction() in
2453 modern systems) in an attempt to fix a rare and obscure crash bug.
2455 15. Narrowed the window for a race in the daemon that could cause it to ignore
2456 SIGCHLD signals. This is not a major problem, because they are used only to
2457 wake it up if nothing else does.
2459 16. A malformed maildirsize file could cause Exim to calculate negative values
2460 for the mailbox size or file count. Odd effects could occur as a result.
2461 The maildirsize information is now recalculated if the size or filecount
2464 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2465 support for a long time. Removed HAVE_SYS_VFS_H.
2467 18. Updated exipick to current release from John Jetmore.
2469 19. Allow an empty sender to be matched against a lookup in an address list.
2470 Previously the only cases considered were a regular expression, or an
2473 20. Exim went into a mad DNS lookup loop when doing a callout where the
2474 host was specified on the transport, if the DNS lookup yielded more than
2477 21. The RFC2047 encoding function was originally intended for short strings
2478 such as real names; it was not keeping to the 75-character limit for
2479 encoded words that the RFC imposes. It now respects the limit, and
2480 generates multiple encoded words if necessary. To be on the safe side, I
2481 have increased the buffer size for the ${rfc2047: expansion operator from
2484 22. Failure to deliver a bounce message always caused it to be frozen, even if
2485 there was an errors_to setting on the router. The errors_to setting is now
2488 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2489 canonical form (fully expanded) before being placed in
2490 $sender_host_address.
2492 24. Updated eximstats to version 1.33
2494 25. Include certificate and key file names in error message when GnuTLS fails
2495 to set them up, because the GnuTLS error message doesn't include the name
2496 of the failing file when there is a problem reading it.
2498 26. Expand error message when OpenSSL has problems setting up cert/key files.
2501 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2502 (this can affect the format of dates).
2504 28. exim_tidydb, when checking for the continued existence of a message for
2505 which it has found a message-specific retry record, was not finding
2506 messages that were in split spool directories. Consequently, it was
2507 deleting retry records that should have stayed in existence.
2509 29. eximstats updated to version 1.35
2510 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2511 1.35 - bugfix such that pie charts by volume are generated correctly
2513 30. The SPA authentication driver was not abandoning authentication and moving
2514 on to the next authenticator when an expansion was forced to fail,
2515 contradicting the general specification for all authenticators. Instead it
2516 was generating a temporary error. It now behaves as specified.
2518 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2519 (the order specifies the preference for clients). The order is now AES256,
2520 AES128, 3DES, ARCFOUR128.
2522 31. Small patch to Sieve code - explicitly set From: when generating an
2525 32. Exim crashed if a remote delivery caused a very long error message to be
2526 recorded - for instance if somebody sent an entire SpamAssassin report back
2527 as a large number of 550 error lines. This bug was coincidentally fixed by
2528 increasing the size of one of Exim's internal buffers (big_buffer) that
2529 happened as part of the Exiscan merge. However, to be on the safe side, I
2530 have made the code more robust (and fixed the comments that describe what
2533 33. Some experimental protocols are using DNS PTR records for new purposes. The
2534 keys for these records are domain names, not reversed IP addresses. The
2535 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2536 leaves it alone. Component reversal etc. now happens only for IP addresses.
2539 34. The host_aton() function is supposed to be passed a string that is known
2540 to be a valid IP address. However, in the case of IPv6 addresses, it was
2541 not checking this. This is a hostage to fortune. Exim now panics and dies
2542 if the condition is not met. A case was found where this could be provoked
2543 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2544 components; fortuitously, this particular loophole had already been fixed
2545 by change 4.50/55 or 4.44/33 above.
2547 If there are any other similar loopholes, the new check in host_aton()
2548 itself should stop them being exploited. The report I received stated that
2549 data on the command line could provoke the exploit when Exim was running as
2550 exim, but did not say which command line option was involved. All I could
2551 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2552 running as the user.
2555 35. There was a buffer overflow vulnerability in the SPA authentication code
2556 (which came originally from the Samba project). I have added a test to the
2557 spa_base64_to_bits() function which I hope fixes it.
2560 36. The daemon start-up calls getloadavg() while still root for those OS that
2561 need the first call to be done as root, but it missed one case: when
2562 deliver_queue_load_max is set with deliver_drop_privilege. This is
2563 necessary for the benefit of the queue runner, because there is no re-exec
2564 when deliver_drop_privilege is set.
2566 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2567 was in use, was not putting the data itself into the right store pool;
2568 consequently, it could be overwritten for a subsequent message in the same
2569 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2572 38. Sometimes the final signoff response after QUIT could fail to get
2573 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2574 < 0 before doing a fflush(). This bug looks as though it goes back to the
2575 introduction of TLS in release 3.20, but "sometimes" must have been rare
2576 because the tests only now provoked it.
2582 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2583 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2584 Now it returns an integer. A number of calls were still expecting a T/F
2585 return. Fortuitously, in all cases, the tests worked in OK situations,
2586 which is the norm. However, things would have gone wrong on any write
2587 failures on the smtp file descriptor. This function is used when sending
2588 messages over SMTP and also when doing verify callouts.
2590 2. When Exim is called to do synchronous delivery of a locally submitted
2591 message (the -odf or -odi options), it no longer closes stderr before doing
2594 3. Implemented the mua_wrapper option.
2596 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2598 5. Implemented the functions header_remove(), header_testname(),
2599 header_add_at_position(), and receive_remove_recipient(), and exported them
2602 6. If an ACL "warn" statement specified the addition of headers, Exim already
2603 inserted X-ACL-Warn: at the start if there was no header name. However, it
2604 was not making this test for the second and subsequent header lines if
2605 there were newlines in the string. This meant that an invalid header could
2606 be inserted if Exim was badly configured.
2608 7. Allow an ACL "warn" statement to add header lines at the start or after all
2609 the Received: headers, as well as at the end.
2611 8. Added the rcpt_4xx retry error code.
2613 9. Added postmaster_mailfrom=xxx to callout verification option.
2615 10. Added mailfrom=xxxx to the callout verification option, for verify=
2618 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2619 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2621 12. Inserted some casts to stop certain compilers warning when using pointer
2622 differences as field lengths or precisions in printf-type calls (mostly
2623 affecting debugging statements).
2625 13. Added optional readline() support for -be (dynamically loaded).
2627 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2628 same clock tick as a message's arrival, so that its received time was the
2629 same as the "first fail" time on the retry record, and that message
2630 remained on the queue past the ultimate address timeout, every queue runner
2631 would try a delivery (because it was past the ultimate address timeout) but
2632 after another failure, the ultimate address timeout, which should have then
2633 bounced the address, did not kick in. This was a "< instead of <=" error;
2634 in most cases the first failure would have been in the next clock tick
2635 after the received time, and all would be well.
2637 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2638 being recognized when the domain list was tested by the match_domain
2639 condition in an expansion string.
2641 16. Added the ${str2b64: operator.
2643 17. Exim was always calling setrlimit() to set a large limit for the number of
2644 processes, without checking whether the existing limit was already
2645 adequate. (It did check for the limit on file descriptors.) Furthermore,
2646 errors from getrlimit() and setrlimit() were being ignored. Now they are
2647 logged to the main and panic logs, but Exim does carry on, to try to do its
2648 job under whatever limits there are.
2650 18. Imported PCRE 5.0.
2652 19. Trivial typo in log message " temporarily refused connection" (the leading
2655 20. If the log selector return_path_on_delivery was set and an address was
2656 redirected to /dev/null, the delivery process crashed because it assumed
2657 that a return path would always be set for a "successful" delivery. In this
2658 case, the whole delivery is bypassed as an optimization, and therefore no
2661 21. Internal re-arrangement: the function for sending a challenge and reading
2662 a response while authentication was assuming a zero-terminated challenge
2663 string. It's now changed to take a pointer and a length, to allow for
2664 binary data in such strings.
2666 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2668 23. Exim was not respecting finduser_retries when seeking the login of the
2669 uid under which it was called; it was always trying 10 times. (The default
2670 setting of finduser_retries is zero.) Also, it was sleeping after the final
2671 failure, which is pointless.
2673 24. Implemented tls_on_connect_ports.
2675 25. Implemented acl_smtp_predata.
2677 26. If the domain in control=submission is set empty, Exim assumes that the
2678 authenticated id is a complete email address when it generates From: or
2679 Sender: header lines.
2681 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2682 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2683 chown and chgrp in /bin and hostname in /usr/bin.
2685 28. Exim was keeping the "process log" file open after each use, just as it
2686 does for the main log. This opens the possibility of it remaining open for
2687 long periods when the USR1 signal hits a daemon. Occasional processlog
2688 errors were reported, that could have been caused by this. Anyway, it seems
2689 much more sensible not to leave this file open at all, so that is what now
2692 29. The long-running daemon process does not normally write to the log once it
2693 has entered its main loop, and it closes the log before doing so. This is
2694 so that log files can straightforwardly be renamed and moved. However,
2695 there are a couple of unusual error situations where the daemon does write
2696 log entries, and I had neglected to close the log afterwards.
2698 30. The text of an SMTP error response that was received during a remote
2699 delivery was being truncated at 512 bytes. This is too short for some of
2700 the long messages that one sometimes sees. I've increased the limit to
2703 31. It is now possible to make retry rules that apply only when a message has a
2704 specific sender, in particular, an empty sender.
2706 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2707 it possible to be selective about when SMTP synchronization is enforced.
2709 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2711 32. Implemented hosts_connection_nolog.
2713 33. Added an ACL for QUIT.
2715 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2718 35. Added mailbox_size and mailbox_filecount to appendfile.
2720 36. Added control = no_multiline_responses to ACLs.
2722 37. There was a bug in the logic of the code that waits for the clock to tick
2723 in the case where the clock went backwards by a substantial amount such
2724 that the microsecond fraction of "now" was more than the microsecond
2725 fraction of "then" (but the whole seconds number was less).
2727 38. Added support for the libradius Radius client library this is found on
2728 FreeBSD (previously only the radiusclient library was supported).
2734 1. When certain lookups returned multiple values in the form name=value, the
2735 quoting of the values was not always being done properly. Specifically:
2736 (a) If the value started with a double quote, but contained no whitespace,
2738 (b) If the value contained whitespace other than a space character (i.e.
2739 tabs or newlines or carriage returns) it was not quoted.
2740 This fix has been applied to the mysql and pgsql lookups by writing a
2741 separate quoting function and calling it from the lookup code. The fix
2742 should probably also be applied to nisplus, ibase and oracle lookups, but
2743 since I cannot test any of those, I have not disturbed their existing code.
2745 2. A hit in the callout cache for a specific address caused a log line with no
2746 reason for rejecting RCPT. Now it says "Previous (cached) callout
2747 verification failure".
2749 3. There was an off-by-one bug in the queryprogram router. An over-long
2750 return line was truncated at 256 instead of 255 characters, thereby
2751 overflowing its buffer with the terminating zero. As well as fixing this, I
2752 have increased the buffer size to 1024 (and made a note to document this).
2754 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2755 when Exim is waiting for an SMTP response from a remote server, Exim
2756 restarts its select() call on the socket, thereby resetting its timeout.
2757 This is not a problem when such interrupts are rare. Somebody set up a cron
2758 job to run exiwhat every 2 minutes, which is less than the normal select()
2759 timeout (5 or 10 minutes). This meant that the select() timeout never
2760 kicked in because it was always reset. I have fixed this by comparing the
2761 time when an interrupt arrives with the time at the start of the first call
2762 to select(). If more time than the timeout has elapsed, the interrupt is
2763 treated as a timeout.
2765 5. Some internal re-factoring in preparation for the addition of Sieve
2766 extensions (by MH). In particular, the "personal" test is moved to a
2767 separate function, and given an option for scanning Cc: and Bcc: (which is
2768 not set for Exim filters).
2770 6. When Exim created an email address using the login of the caller as the
2771 local part (e.g. when creating a From: or Sender: header line), it was not
2772 quoting the local part when it contained special characters such as @.
2774 7. Installed new OpenBSD configuration files.
2776 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2777 try to make them clearer.
2779 9. Callout options, other than the timeout value, were being ignored when
2780 verifying sender addresses in header lines. For example, when using
2782 verify = header_sender/callout=no_cache
2784 the cache was (incorrectly) being used.
2786 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2787 only the Cygwin environment.
2789 11. When return_path_on_delivery was set as a log selector, if different remote
2790 addresses in the same message used different return paths and parallel
2791 remote delivery occurred, the wrong values would sometimes be logged.
2792 (Whenever a remote delivery process finished, the return path value from
2793 the most recently started remote delivery process was logged.)
2795 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2796 lines when AUTH and/or TLS are in use. This is the "received protocol"
2797 field. Exim used to use "asmtp" for authenticated SMTP, without any
2798 indication (in the protocol name) for TLS use. Now it follows the RFC and
2799 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2800 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2801 names appear in log lines as well as in Received: header lines.
2803 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2804 extensions, and comparison tests, and to fix some bugs.
2806 14. Changes to the "personal" filter test:
2808 (1) The test was buggy in that it was just doing the equivalent of
2809 "contains" tests on header lines. For example, if a user's address was
2810 anne@some.where, the "personal" test would incorrectly be true for
2812 To: susanne@some.where
2814 This test is now done by extracting each address from the header in turn,
2815 and checking the entire address. Other tests that are part of "personal"
2816 are now done using regular expressions (for example, to check local parts
2817 of addresses in From: header lines).
2819 (2) The list of non-personal local parts in From: addresses has been
2820 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2821 taken from the Sieve specification recommendations.
2823 (3) If the message contains any header line starting with "List-" it is
2824 treated as non-personal.
2826 (4) The test for "circular" in the Subject: header line has been removed
2827 because it now seems ill-conceived.
2829 15. Minor typos in src/EDITME comments corrected.
2831 16. Installed latest exipick from John Jetmore.
2833 17. If headers_add on a router specified a text string that was too long for
2834 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2835 of string_sprintf() is now avoided.
2837 18. $message_body_size was not set (it was always zero) when running the DATA
2838 ACL and the local_scan() function.
2840 19. For the "mail" command in an Exim filter, no default was being set for
2841 the once_repeat time, causing a random time value to be used if "once" was
2842 specified. (If the value happened to be <= 0, no repeat happened.) The
2843 default is now 0s, meaning "never repeat". The "vacation" command was OK
2844 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2845 (I found it when inspecting the code).
2847 20. There is now an overall timeout for performing a callout verification. It
2848 defaults to 4 times the callout timeout, which applies to individual SMTP
2849 commands during the callout. The overall timeout applies when there is more
2850 than one host that can be tried. The timeout is checked before trying the
2851 next host. This prevents very long delays if there are a large number of
2852 hosts and all are timing out (e.g. when the network connections are timing
2853 out). The value of the overall timeout can be changed by specifying an
2854 additional sub-option for "callout", called "maxwait". For example:
2856 verify = sender/callout=5s,maxwait=20s
2858 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2859 to the end before writing, but this should make it even safer).
2861 22. Exim was forgetting that it had advertised PIPELINING for the second and
2862 subsequent messages on an SMTP connection. It was also not resetting its
2863 memory on STARTTLS and an internal HELO.
2865 23. When Exim logs an SMTP synchronization error within a session, it now
2866 records whether PIPELINING has been advertised or not.
2868 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2869 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2870 rather than long int.
2872 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2874 26. Added the never_mail option to autoreply.
2880 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2881 crash if the getsockname() call failed; this can happen if a connection is
2882 closed very soon after it is established. The problem was simply in the
2883 order in which certain operations were done, causing Exim to try to write
2884 to the SMTP stream before it had set up the file descriptor. The bug has
2885 been fixed by making things happen in the correct order.
2891 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2892 before the connection was closed, thus losing the rejection response.
2894 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2895 some early Solaris releases, but causes trouble in current releases where
2896 socklen_t is defined.
2898 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2901 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2904 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2905 that releasing the top of it at the end releases what was used for sub-
2906 expansions (unless the block got too big). However, discard this block if
2907 the first thing is a variable or header, so that we can use its block when
2908 it is dynamic (useful for very large $message_headers, for example).
2910 6. Lookups now cache *every* query, not just the most recent. A new, separate
2911 store pool is used for this. It can be recovered when all lookup caches are
2912 flushed. Lookups now release memory at the end of their result strings.
2913 This has involved some general refactoring of the lookup sources.
2915 7. Some code has been added to the store_xxx() functions to reduce the amount
2916 of flapping under certain conditions.
2918 8. log_incoming_interface used to affect only the <= reception log lines. Now
2919 it causes the local interface and port to be added to several more SMTP log
2920 lines, for example "SMTP connection from", and rejection lines.
2922 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2924 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2926 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2927 could be overwritten at the end of the current message (or the start of a
2928 new message if it was set in a HELO ACL). The value is now preserved for
2929 the duration of the SMTP connection.
2931 12. If a transport had a headers_rewrite setting, and a matching header line
2932 contained an unqualified address, that address was qualified, even if it
2933 did not match any rewriting rules. The underlying bug was that the values
2934 of the flags that permit the existence of unqualified sender and recipient
2935 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2936 non-local messages, and by -bnq for local messages) were not being
2937 preserved with the message after it was received.
2939 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2940 "next input=" as part of the text comprising the host identity instead of
2941 the correct text. The code was using the same buffer for two different
2942 strings. However, depending on which order the printing function evaluated
2943 its arguments, the bug did not always show up. Under Linux, for example, my
2944 test suite worked just fine.
2946 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2947 doesn't work with some older versions of Perl. It has been changed to "my",
2948 which in any case is probably the better facility to use.
2950 15. A really picky compiler found some instances of statements for creating
2951 error messages that either had too many or two few arguments for the format
2954 16. The size of the buffer for calls to the DNS resolver has been increased
2955 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2956 for addresses that have a lot of PTR records. This alleviates a problem; it
2957 does not fully solve it.
2959 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2960 buffer now truncates the list and logs the incident, which is the same
2961 action as happens when Exim is looking up a host name and its aliases.
2962 Previously in this situation something unpredictable would happen;
2963 sometimes it was "internal error: store_reset failed".
2965 18. If a server dropped the connection unexpectedly when an Exim client was
2966 using GnuTLS and trying to read a response, the client delivery process
2967 crashed while trying to generate an error log message.
2969 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2970 string, for example:
2972 warn message = H1: something\nH2: something
2974 the text was added as a single header line from Exim's point of view
2975 though it ended up OK in the delivered message. However, searching for the
2976 second and subsequent header lines using $h_h2: did not work. This has been
2977 fixed. Similarly, if a system filter added multiple headers in this way,
2978 the routers could not see them.
2980 20. Expanded the error message when iplsearch is called with an invalid key to
2981 suggest using net-iplsearch in a host list.
2983 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2984 ACLs are no longer actually imposed (and a message to that effect is
2987 22. If a "gecos" field in a passwd entry contained escaped characters, in
2988 particular, if it contained a \" sequence, Exim got it wrong when building
2989 a From: or a Sender: header from that name. A second bug also caused
2990 incorrect handling when an unquoted " was present following a character
2991 that needed quoting.
2993 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2994 was not being matched caselessly.
2996 24. Arranged for all hyphens in the exim.8 source to be escaped with
2999 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3000 properly. Recipient callout cache records were still being keyed to include
3001 the sender, even when use_sender was set false. This led to far more
3002 callouts that were necessary. The sender is no longer included in the key
3003 when use_sender is false.
3005 26. Added "control = submission" modifier to ACLs.
3007 27. Added the ${base62d: operator to decode base 62 numbers.
3009 28. dnsdb lookups can now access SRV records.
3011 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3012 the configuration file.
3014 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3015 (-v) mode. This makes the output for a verbose queue run more intelligible.
3017 31. Added a use_postmaster feature to recipient callouts.
3019 32. Added the $body_zerocount variable, containing the number of binary zero
3020 bytes in the message body.
3022 33. The time of last modification of the "new" subdirectory is now used as the
3023 "mailbox time last read" when there is a quota error for a maildir
3026 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3028 35. Added +ignore_unknown as a special item in host lists.
3030 36. Code for decoding IPv6 addresses in host lists is now included, even if
3031 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3032 address was recognized as an IP address, but was then not correctly decoded
3033 into binary, causing unexpected and incorrect effects when compared with
3040 1. Very minor rewording of debugging text in manualroute to say "list of
3041 hosts" instead of "hostlist".
3043 2. If verify=header_syntax was set, and a header line with an unqualified
3044 address (no domain) and a large number of spaces between the end of the
3045 name and the colon was received, the reception process suffered a buffer
3046 overflow, and (when I tested it) crashed. This was caused by some obsolete
3047 code that should have been removed. The fix is to remove it!
3049 3. When running in the test harness, delay a bit after writing a bounce
3050 message to get a bit more predictability in the log output.
3052 4. Added a call to search_tidyup() just before forking a reception process. In
3053 theory, someone could use a lookup in the expansion of smtp_accept_max_
3054 per_host which, without the tidyup, could leave open a database connection.
3056 5. Added the variables $recipient_data and $sender_data which get set from a
3057 lookup success in an ACL "recipients" or "senders" condition, or a router
3058 "senders" option, similar to $domain_data and $local_part_data.
3060 6. Moved the writing of debug_print from before to after the "senders" test
3063 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3064 problems for message scanning, either using a data ACL, or using
3065 local_scan() because the Received: header was not generated till after they
3066 were called (in order to set the time as the time of reception completion).
3067 I have revised the way this works. The header is now generated after the
3068 body is received, but before the ACL or local_scan() are called. After they
3069 are run, the timestamp in the header is updated.
3075 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3076 before starting a queue runner without re-exec. This happened only when
3077 deliver_drop_privilege was set or when the Exim user was set to root. The
3078 effect of the bug was that timeouts during subsequent deliveries caused
3079 crashes instead of being properly handled. The handler is now left at its
3080 default (and expected) setting.
3082 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3083 message, again when deliver_drop_privilege is set or Exim is run as root.
3084 The bug described in (1) was not present in this case, but the tidying up
3085 of the other signals was missing. I have made the two cases consistent.
3087 3. The ignore_target_hosts setting on a manualroute router was being ignored
3088 for hosts that were looked up using the /MX notation.
3090 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3093 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3094 operated on the sender address. After changing the $sender_address to <>
3095 for the sender address verify, Exim was re-instated it as the original
3096 (before rewriting) address, but remembering that it had rewritten it, so it
3097 wasn't rewriting it again. This bug also had the effect of breaking the
3098 sender address verification caching when the sender address was rewritten.
3100 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3101 This has been changed so that if the ip literal address matches
3102 ignore_target_hosts, the router declines.
3104 7. Added expansion conditions match_domain, match_address, and match_local_
3105 part (NOT match_host).
3107 8. The placeholder for the Received: header didn't have a length field set.
3109 9. Added code to Exim itself and to exim_lock to test for a specific race
3110 condition that could lead to file corruption when using MBX delivery. The
3111 issue is with the lockfile that is created in /tmp. If this file is removed
3112 after a process has opened it but before that process has acquired a lock,
3113 there is the potential for a second process to recreate the file and also
3114 acquire a lock. This could lead to two Exim processes writing to the file
3115 at the same time. The added code performs the same test as UW imapd; it
3116 checks after acquiring the lock that its file descriptor still refers to
3117 the same named file.
3119 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3120 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3123 11. Added the smtp_active_hostname option. If used, this will typically be made
3124 to depend on the incoming interface address. Because $interface_address is
3125 not set up until the daemon has forked a reception process, error responses
3126 that can happen earlier (such as "too many connections") no longer contain
3129 12. If an expansion in a condition on a "warn" statement fails because a lookup
3130 defers, the "warn" statement is abandoned, and the next ACL statement is
3131 processed. Previously this caused the whole ACL to be aborted.
3133 13. Added the iplsearch lookup type.
3135 14. Added ident_timeout as a log selector.
3137 15. Added tls_certificate_verified as a log selector.
3139 16. Added a global option tls_require_ciphers (compare the smtp transport
3140 option of the same name). This controls incoming TLS connections.
3142 17. I finally figured out how to make tls_require_ciphers do a similar thing
3143 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3144 before starting the TLS session.
3146 18. Tabs are now shown as \t in -bP output.
3148 19. If the log selector return_path_on_delivery was set, Exim crashed when
3149 bouncing a message because it had too many Received: header lines.
3151 20. If two routers both had headers_remove settings, and the first one included
3152 a superfluous trailing colon, the final name in the first list and the
3153 first name in the second list were incorrectly joined into one item (with a
3154 colon in the middle).
3160 1. Added -C and -D options to the exinext utility, mainly to make it easier
3161 to include in the automated testing, but these could be helpful when
3162 multiple configurations are in use.
3164 2. The exinext utility was not formatting the output nicely when there was
3165 an alternate port involved in the retry record key, nor when there was a
3166 message id as well (for retries that were specific to a specific message
3167 and a specific host). It was also confused by IPv6 addresses, because of
3168 the additional colons they contain. I have fixed the IPv4 problem, and
3169 patched it up to do a reasonable job for IPv6.
3171 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3172 delivery, the log line now contains "pipelined" if PIPELINING was used.
3174 4. An SMTP transport process used to panic and die if the bind() call to set
3175 an explicit outgoing interface failed. This has been changed; it is now
3176 treated in the same way as a connect() failure.
3178 5. A reference to $sender_host_name in the part of a conditional expansion
3179 that was being skipped was still causing a DNS lookup. This no longer
3182 6. The def: expansion condition was not recognizing references to header lines
3183 that used bh_ and bheader_.
3185 7. Added the _cache feature to named lists.
3187 8. The code for checking quota_filecount in the appendfile transport was
3188 allowing one more file than it should have been.
3190 9. For compatibility with Sendmail, the command line option
3198 and sets the incoming protocol and host name (for trusted callers). The
3199 host name and its colon can be omitted when only the protocol is to be set.
3200 Note the Exim already has two private options, -pd and -ps, that refer to
3201 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3202 "s", but I don't think that's a major issue.
3204 10. A number of refactoring changes to the code, none of which should affect
3207 (a) The number of logging options was getting close to filling up the
3208 32-bit word that was used as a bit map. I have split them into two classes:
3209 those that are passed in the argument to log_write(), and those that are
3210 only ever tested independently outside of that function. These are now in
3211 separate 32-bit words, so there is plenty of room for expansion again.
3212 There is no change in the user interface or the logging behaviour.
3214 (b) When building, for example, log lines, the code previously used a
3215 macro that called string_cat() twice, in order to add two strings. This is
3216 not really sufficiently general. Furthermore, there was one instance where
3217 it was actually wrong because one of the argument was used twice, and in
3218 one call a function was used. (As it happened, calling the function twice
3219 did not affect the overall behaviour.) The macro has been replaced by a
3220 function that can join an arbitrary number of extra strings onto a growing
3223 (c) The code for expansion conditions now uses a table and a binary chop
3224 instead of a serial search (which was left over from when there were very
3225 few conditions). Also, it now recognizes conditions like "pam" even when
3226 the relevant support is not compiled in: a suitably worded error message is
3227 given if an attempt is made to use such a condition.
3229 11. Added ${time_interval:xxxxx}.
3231 12. A bug was causing one of the ddress fields not to be passed back correctly
3232 from remote delivery subprocesses. The field in question was not being
3233 subsequently used, so this caused to problems in practice.
3235 13. Added new log selectors queue_time and deliver_time.
3237 14. Might have fixed a bug in maildirsizefile handling that threw up
3238 "unexpected character" debug warnings, and recalculated the data
3239 unnecessarily. In any case, I expanded the warning message to give more
3242 15. Added the message "Restricted characters in address" to the statements in
3243 the default ACL that block characters like @ and % in local parts.
3245 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3246 Three changes have been made:
3248 (a) There was a serious bug; a negative response to MAIL caused the whole
3249 recipient domain to be cached as invalid, thereby blocking all messages
3250 to all local parts at the same domain, from all senders. This bug has
3251 been fixed. The domain is no longer cached after a negative response to
3252 MAIL if the sender used is not empty.
3254 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3257 (c) A new callout option, "use_sender" has been added for people who want
3258 the modified behaviour.
3264 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3267 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3268 indeed breaks things for older releases.
3270 3. Added additional logging to the case where there is a problem reading data
3271 from a filter that is running in a subprocess using a pipe, in order to
3272 try to track down a specific problem.
3274 4. Testing facility fudge: when running in the test harness and attempting
3275 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3276 getting "No route to host". Convert this to a timeout.
3278 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3281 6. Some OS don't have socklen_t but use size_t instead. This affects the
3282 fifth argument of getsockopt() amongst other things. This is now
3283 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3284 can be set for individual OS. I have set it for SunOS5, OSF1, and
3285 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3286 some earlier ones do not.
3288 7. Change 4.30/15 was not doing the test caselessly.
3290 8. The standard form for an IPv6 address literal was being rejected by address
3291 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3292 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3293 this, as well as the form without the "IPv6" on the front (but only when
3294 address literals are enabled, of course).
3296 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3298 10. Exim crashed if a message with an empty sender address specified by -f
3299 encountered a router with an errors_to setting. This could be provoked only
3300 by a command such as
3304 where an empty string was supplied; "<>" did not hit this bug.
3306 11. Installed PCRE release 4.5.
3308 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3309 remained set. It is now erased.
3311 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3312 times from message ids (which are base 36 rather than the normal 62).
3314 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3315 were being counted as actual protocol errors, and logged if the log
3316 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3317 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3318 and DATA following a set of rejected RCPTs do not count as protocol errors.
3319 In other words, Exim assumes they were pipelined, though this may not
3320 actually be the case. Of course, in all cases the client gets an
3321 appropriate error code.
3323 15. If a lookup fails in an ACL condition, a message about the failure may
3324 be available; it is used if testing the ACL cannot continue, because most
3325 such messages specify what the cause of the deferral is. However, some
3326 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3327 that caused an old message to be retained and used if a later statement
3328 caused a defer, replacing the real cause of the deferral.
3330 16. If an IP address had so many PTR records that the DNS lookup buffer
3331 was not large enough to hold them, Exim could crash while trying to process
3332 the truncated data. It now detects and logs this case.
3334 17. Further to 4.21/58, another change has been made: if (and only if) the
3335 first line of a message (the first header line) ends with CRLF, a bare LF
3336 in a subsequent header line has a space inserted after it, so as not to
3337 terminate the header.
3339 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3340 unnecessarily, used atoi() instead of strtol(), and didn't check the
3341 termination when getting file sizes from file names by regex.
3343 19. Completely re-implemented the support for maildirsize files, in the light
3344 of a number of problems with the previous contributed implementation
3345 (4.30/29). In particular:
3347 . If the quota is zero, the maildirsize file is maintained, but no quota is
3350 . If the maildir directory does not exist, it is created before any attempt
3351 to write a maildirsize file.
3353 . The quota value in the file is just a cache; if the quota is changed in
3354 the transport, the new value overrides.
3356 . A regular expression is available for excluding directories from the
3359 20. The autoreply transport checks the characters in options that define the
3360 message's headers; it allows continued headers, but it was checking with
3361 isspace() after an embedded newline instead of explicitly looking for a
3364 21. If all the "regular" hosts to which an address was routed had passed their
3365 expiry times, and had not reached their retry times, the address was
3366 bounced, even if fallback hosts were defined. Now Exim should go on to try
3369 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3370 equivalent code in the SMTP transport. Some hosts send humungous responses
3371 to HELO/EHLO, more than 1024 it seems.
3373 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3374 gives compiler warnings in some environments. I've now done it "properly",
3377 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3378 (because of problems with the built-in one) was declared to return uschar *
3379 instead of char *, causing compiler failure.
3381 25. Fixed a file descriptor leak when processing alias/forward files.
3383 26. Fixed a minor format string issue in dbfn.c.
3385 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3387 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3388 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3391 29. When a sender address is verified, it is cached, to save repeating the test
3392 when there is more than one recipient in a message. However, when the
3393 verification involves a callout, it is possible for different callout
3394 options to be set for different recipients. It is too complicated to keep
3395 track of this in the cache, so now Exim always runs a verification when a
3396 callout is required, relying on the callout cache for the optimization.
3397 The overhead is duplication of the address routing, but this should not be
3400 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3401 to be verified with callout=postmaster, and the main callout worked but the
3402 postmaster check failed, the verification correctly failed. However, if a
3403 subsequent RCPT command asked for sender verification *without* the
3404 postmaster check, incorrect caching caused this verification also to fail,
3407 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3408 it was not caching the DNS options (qualify_single, search_parents) that
3409 were used when the lookup failed. A subsequent lookup with different
3410 options therefore always gave the same answer, though there were cases
3411 where it should not have. (Example: a "domains = !$mx_any" option on a
3412 dnslookup router: the "domains" option is always processed without any
3413 widening, but the router might have qualify_single set.) Now Exim uses the
3414 cached value only when the same options are set.
3416 32. Added John Jetmore's "exipick" utility to the distribution.
3418 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3419 than a timeout (e.g. a certificate is required, and is not provided), an
3420 Exim server now closes the connection immediately. Previously it waited for
3421 the client to close - but if the client is SSL, it seems that they each
3422 wait for each other, leading to a delay before one of them times out.
3424 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3425 maintained 0.8.x compatibility because I don't think many are using it, and
3426 it is clearly obsolete.
3428 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3431 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3432 client certificate was expired. A simple patch fixes this, though I don't
3433 understand the full logic of why the verify callback is called multiple
3436 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3437 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3438 which causes problems with some clients (such as the Certicom SSL Plus
3439 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3440 disables the coutermeasure allowing Eudora to connect."
3442 38. Exim was not checking that a write() to a log file succeeded. This could
3443 lead to Bad Things if a log got too big, in particular if it hit a file
3444 size limit. Exim now panics and dies if it cannot write to a log file, just
3445 as it does if it cannot open a log file.
3447 39. Modified OS/Makefile-Linux so that it now contains
3449 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3451 The two -D definitions ensure that Exim is compiled with large file
3452 support, which makes it possible to handle log files that are bigger than
3455 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3456 instance) a domain was checked against a named list that involved a lookup,
3457 causing $domain_data to be set, then another domain was checked against the
3458 same list, then the first domain was re-checked, the value of $domain_data
3459 after the final check could be wrong. In particular, if the second check
3460 failed, it could be set empty. This bug probably also applied to
3463 41. The strip_trailing_dot option was not being applied to the address given
3464 with the -f command-line option.
3466 42. The code for reading a message's header from the spool was incrementing
3467 $received_count, but never initializing it. This meant that the value was
3468 incorrect (doubled) while delivering a message in the same process in which
3469 it was received. In the most common configuration of Exim, this never
3470 happens - a fresh exec is done - but it can happen when
3471 deliver_drop_privilege is set.
3473 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3474 it now includes up to 150 characters of the unexpected data in the log
3477 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3478 and building data strings. The size of both of these buffers was 10 000
3479 bytes - far larger than anybody would *ever* want, thought I. Needless to
3480 say, somebody hit the limit. I have increased the maximum line length to
3481 20 000 and the maximum data length of concatenated lines to 100 000. I have
3482 also fixed two bugs, because there was no checking on these buffers. Tsk,
3483 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3484 buffer is too small.
3486 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3487 lsearch lookups. Now it does.
3489 46. When parsing a route_list item in a manualroute router, a fixed-length
3490 buffer was used for the list of hosts. I made this 1024 bytes long,
3491 thinking that nobody would ever have a list of hosts that long. Wrong.
3492 Somebody had a whole pile of complicated expansion conditions, and the
3493 string was silently truncated, leading to an expansion error. It turns out
3494 that it is easier to change to an unlimited length (owing to other changes
3495 that have happened since this code was originally written) than to build
3496 structure for giving a limitation error. The length of the item that
3497 expands into the list of hosts is now unlimited.
3499 47. The lsearch lookup could not handle data where the length of text line was
3500 more than 4095 characters. Such lines were truncated, leading to shortened
3501 data being returned. It should now handle lines of any length.
3503 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3504 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3507 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3508 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3509 no longer happens in comment lines. A list of replacements is now placed
3510 at the head of all of the source files, except those whose only change is
3511 to replace PERL_COMMAND in the very first #! line.
3513 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3514 messages on the queue, with a bottom-up merge sort, using code contributed
3515 by Michael Haardt. This should make operations like -bp somewhat faster on
3516 large queues. It won't affect queue runners, except when queue_run_in_order
3519 51. Installed eximstats 1.31 in the distribution.
3521 52. Added support for SRV lookups to the dnslookup router.
3523 53. If an ACL referred to $message_body or $message_body_end, the value was not
3524 reset for any messages that followed in the same SMTP session.
3526 54. The store-handling optimization for building very long strings was not
3527 differentiating between the different store pools. I don't think this
3528 actually made any difference in practice, but I've tidied it.
3530 55. While running the routers to verify a sender address, $sender_address
3531 was still set to the sender address. This is wrong, because when routing to
3532 send a bounce to the sender, it would be empty. Therefore, I have changed
3533 it so that, while verifying a sender address, $sender_address is set to <>.
3534 (There is no change to what happens when verifying a recipient address.)
3536 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3537 target A or AAAA records (if not already returned) without resetting the
3538 qualify_single or search_parents options of the DNS resolver. These are
3539 inappropriate in this case because the targets of MX and SRV records must
3540 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3541 target that, when qualified, matched something in the local domain. These
3542 two options are now turned off when doing these lookups.
3544 57. It seems that at least some releases of Reiserfs (which does not have the
3545 concept of a fixed number of inodes) returns zero and not -1 for the
3546 number of available inodes. This interacted badly with check_spool_inodes,
3547 which assumed that -1 was the "no such thing" setting. What I have done is
3548 to check that the total number of inodes is greater than zero before doing
3549 the test of how many are available.
3551 58. When a "warn" ACL statement has a log_message modifier, the message is
3552 remembered, and not repeated. This is to avoid a lot of repetition when a
3553 message has many recipients that cause the same warning to be written.
3554 Howewer, Exim was preserving the list of already written lines for an
3555 entire SMTP session, which doesn't seem right. The memory is now reset if a
3556 new message is started.
3558 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3559 debugging output unless log_rewrite was also set.
3561 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3562 of (int)(handle) when we know that handle contains (void *)(-1).
3564 61. The Exim daemon panic-logs an error return when it closes the incoming
3565 connection. However "connection reset by peer" seems to be common, and
3566 isn't really an error worthy of noting specially, so that particular error
3569 62. When Exim is trying to find all the local interfaces, it used to panic and
3570 die if the ioctl to get the interface flags failed. However, it seems that
3571 on at least one OS (Solaris 9) it is possible to have an interface that is
3572 included in the list of interfaces, but for which you get a failure error
3573 for this call. This happens when the interface is not "plumbed" into a
3574 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3575 failure of the "get flags" call assumes that the interface is down.
3577 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3578 makes life easier for people who are doing arithmetic on fields extracted
3579 from dates, where you often get leading zeros that should not be
3580 interpreted as octal.
3582 64. Added qualify_domain to the redirect router, to override the global
3585 65. If a pathologically long header line contained very many addresses (the
3586 report of this problem mentioned 10 000) and each of them was rewritten,
3587 Exim could use up a very large amount of memory. (It kept on making new
3588 copies of the header line as it rewrote, and never released the old ones.)
3589 At the expense of a bit more processing, the header rewriting function has
3590 been changed so that it no longer eats memory in this way.
3592 66. The generation of the Received: header has been moved from the time that a
3593 message starts to be received, to the time that it finishes. The timestamp
3594 in the Received: header should now be very close to that of the <= log
3595 line. There are two side-effects of this change:
3597 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3598 logged header lines no longer include the local Received: line, because
3599 it has not yet been created. The same applies to a copy of the message
3600 that is returned to a non-SMTP sender when a message is rejected.
3602 (b) When a filter file is tested using -bf, no additional Received: header
3603 is added to the test message. After some thought, I decided that this
3606 This change does not affect the value of $received_for. It is still set
3607 after address rewriting, but before local_scan() is called.
3609 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3611 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3612 gave an unhelpful panic error message, and a defer error. I have managed to
3613 change this behaviour so that it now rejects any supplied certificate,
3614 which seems right, as the list of acceptable certificates is empty.
3616 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3617 gave an unhelpful defer error. I have not managed to make this reject any
3618 supplied certificates, but the error message it gives is "no certificate
3619 supplied", which is not helpful.
3621 70. exigrep's output now also includes lines that are not associated with any
3622 message, but which match the given pattern. Implemented by a patch from
3623 Martin Sluka, which also tidied up the Perl a bit.
3625 71. Recipient callout verification, like sender verification, was using <> in
3626 the MAIL FROM command. This isn't really the right thing, since the actual
3627 sender may affect whether the remote host accepts the recipient or not. I
3628 have changed it to use the actual sender in the callout; this means that
3629 the cache record is now keyed on a recipient/sender pair, not just the
3630 recipient address. There doesn't seem to be a real danger of callout loops,
3631 since a callout by the remote host to check the sender would use <>.
3632 [SEE ABOVE: changed after hitting problems.]
3634 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3635 temporary errors. However, in the case of such a code being given after
3636 the end of a data transmission (i.e. after ".") Exim was failing to write
3637 a retry record for the message. (Yes, there was some broken host that was
3638 actually sending 8xx at this point.)
3640 73. An unknown lookup type in a host list could cause Exim to panic-die when
3641 the list was checked. (An example that provoked this was putting <; in the
3642 middle of a list instead of at the start.) If this happened during a DATA
3643 ACL check, a -D file could be left lying around. This kind of configuration
3644 error no longer causes Exim to die; instead it causes a defer errror. The
3645 incident is still logged to the main and panic logs.
3647 74. Buglet left over from Exim 3 conversion. The message "too many messages
3648 in one connection" was written to the rejectlog but not the mainlog, except
3649 when address rewriting (yes!) was being logged.
3651 75. Added write_rejectlog option.
3653 76. When a system filter was run not as root (that is, when system_filter_user
3654 was set), the values of the $n variables were not being returned to the
3655 main process; thus, they were not subsequently available in the $sn
3658 77. Added +return_path_on_delivery log selector.
3660 78. A connection timeout was being treated differently from recipients deferred
3661 when testing hosts_max_try with a message that was older than the host's
3662 retry timeout. (The host should not be counted, thus allowing all hosts to
3663 be tried at least once before bouncing.) This may have been the cause of an
3664 occasionally reported bug whereby a message would remain on the queue
3665 longer than the retry timeout, but would be bounced if a delivery was
3666 forced. I say "may" because I never totally pinned down the problem;
3667 setting up timeout/retry tests is difficult. See also the next item.
3669 79. The ultimate address timeout was not being applied to errors that involved
3670 a combination of host plus message (for example, a timeout on a MAIL
3671 command). When an address resolved to a number of possible hosts, and they
3672 were not all tried for each delivery (e.g. because of hosts_max_try), a
3673 message could remain on the queue longer than the retry timeout.
3675 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3678 81. Fixed an obscure SMTP outgoing bug which required at least the following
3679 conditions: (a) there was another message waiting for the same server;
3680 (b) the server returned 5xx to all RCPT commands in the first message so
3681 that the message was not completed; (c) the server dropped the connection
3682 or gave a negative response to the RSET that Exim sends to abort the
3683 transaction. The observed case was a dropped connection after DATA that had
3684 been sent in pipelining mode. That is, the server had advertised PIPELINING
3685 but was not implementing it correctly. The effect of the bug was incorrect
3686 behaviour, such as trying another host, and this could lead to a crash.
3692 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3693 and daemon.c were passed as pointers to ints; they should have been
3694 pointers to socklen_t variables (which are typically unsigned ints).
3696 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3699 3. Fixed a really odd bug that affected only the testing scheme; patching a
3700 certain fixed string in the binary changed the value of another string that
3701 happened to be identical to the end of the original first string.
3703 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3704 name", it returns that address as the IP address. On some operating
3705 systems (e.g. Solaris), it also passes back the IP address string as the
3706 "host name". However, on others (e.g. Linux), it passes back an empty
3707 string. Exim wasn't checking for this, and was changing the host name to an
3708 empty string, assuming it had been canonicized.
3710 5. Although rare, it is permitted to have more than one PTR record for a given
3711 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3712 all the names associated with an address, because they do in Solaris.
3713 However, it seems that they do not in Linux for data that comes from the
3714 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3715 I found this out when I moved to a new Linux workstation and tried to run
3716 the Exim test suite.
3718 To get round this problem I have changed the code so that it now does its
3719 own call to the DNS to look up PTR records when searching for a host name.
3720 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3721 addresses that are only in /etc/hosts are still found.
3723 This behaviour is, however, controlled by an option called host_lookup_
3724 order, which defaults to "bydns:byaddr". If people want to use the other
3725 order, or indeed, just use one or the other means of lookup, they can
3726 specify it in this variable.
3728 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3729 some operating systems, this comes back from gethostbyaddr() as an empty
3730 string, and this is what Exim used to test for. However, it seems that in
3731 other systems, "." is yielded. Exim now tests for this case too.
3733 7. The values of check_spool_space and check_log_space are now held internally
3734 as a number of kilobytes instead of an absolute number of bytes. If a
3735 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3736 kilobyte. This means that much larger values can be stored.
3738 8. Exim monitor: an attempt to get the action menu when not actually pointing
3739 at a message produces an empty menu entitled "No message selected". This
3740 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3741 no entries in it ("Shell widget menu has zero width and/or height"). So I
3742 have added a single, blank menu entry in this case.
3744 9. Added ${quote_local_part.
3746 10. MIME decoding is now applied to the contents of Subject: header lines when
3749 11. Now that a reference to $sender_host_address automatically causes a reverse
3750 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3751 host lookup before query-style lookups in lists that might use this
3752 variable. This has therefore been abolished, and the "net-" prefix is no
3753 longer necessary for query-style lookups.
3755 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3756 have been a typo for LFLAGS, so it has been changed.
3758 13. The install script calls Exim with "-C /dev/null" in order to find the
3759 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3760 to be output. Howeve, since Exim outputs its version number before the
3761 error, it didn't break the script. It just looked ugly. I fixed this by
3762 always allowing "-C /dev/null" if the caller is root.
3764 14. Ignore overlarge ACL variable number when reading spool file - insurance
3765 against a later release with more variables having written the file.
3767 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3768 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3769 this, as well as the form without the "IPv6" on the front.
3771 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3772 OS/Makefile-Darwin file.
3774 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3775 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3777 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3778 variable. However, some OS use ints for inodes. I've added cast to long int
3779 to get rid of the compiler warning.
3781 19. I had forgotten to lock out "/../" in configuration file names when
3782 ALT_CONFIG_PREFIX was set.
3784 20. Routers used for verification do not need to specify transports. However,
3785 if such a router generated a host list, and callout was configured, Exim
3786 crashed, because it could not find a port number from the (non-existent)
3787 transport. It now assumes port 25 in this circumstance.
3789 21. Added the -t option to exigrep.
3791 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3792 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3793 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3794 them is set, LOOKUP_LSEARCH is forced.
3796 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3798 24. Added sender and host information to the "rejected by local_scan()" log
3799 line; previously there was no indication of these.
3801 25. Added .include_if_exists.
3803 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3804 Linux. It turns out that not all file systems support this. Apparently some
3805 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3806 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3807 not supported on the file descriptor, is now ignored when Exim is trying to
3808 sync a directory. This applies only to Linux.
3810 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3812 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3813 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3816 29. Added support for maildirsize files from supplied patch (modified a bit).
3818 30. The use of :fail: followed by an empty string could lead Exim to respond to
3819 sender verification failures with (e.g.):
3821 550 Verification failed for <xxx>
3822 550 Sender verify failed
3824 where the first response line was missing the '-' that indicates it is not
3825 the final line of the response.
3827 31. The loop for finding the name of the user that called Exim had a hardwired
3828 limit of 10; it now uses the value of finduser_retries, which is used for
3829 all other user lookups.
3831 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3834 33. Exim was neglecting to zero errno before one call of strtol() when
3835 expanding a string and expecting an integer value. On some systems this
3836 resulted in spurious "integer overflow" errors. Also, it was casting the
3837 result into an int without checking.
3839 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3840 did not work. The code looks as if it has *never* worked, though it appears
3841 to have been documented since at least releast 1.62. I have made it work.
3843 35. The "timeout_DNS" error in retry rules, also documented since at least
3844 1.62, also never worked. As it isn't clear exactly what this means, and
3845 clearly it isn't a major issue, I have abolished the feature by treating it
3846 as "timeout", and writing a warning to the main and panic logs.
3848 36. The display of retry rules for -brt wasn't always showing the error code
3851 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3852 timeout_connect_A, timeout_connect_MX.
3854 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3855 to the empty sender.
3857 39. The daemon was not analysing the content of -oX till after it had closed
3858 stderr and disconnected from the controlling terminal. This meant that any
3859 syntax errors were only noted on the panic log, and the return code from
3860 the command was 0. By re-arranging the code a little, I've made the
3861 decoding happen first, so such errors now appear on stderr, and the return
3862 code is 1. However, the actual setting up of the sockets still happens in
3863 the disconnected process, so errors there are still only recorded on the
3866 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3867 connections (as happens on some IP stacks) was logged at start up time as
3868 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3869 it from "IPv6 and IPv4", which means that two separate sockets are being
3872 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3873 says whether AF_INET or AF_INET6 was passed as an argument.
3875 42. Exiwhat output was messed up when time zones were included in log
3878 43. Exiwhat now gives more information about the daemon's listening ports,
3879 and whether -tls-on-connect was used.
3881 44. The "port" option of the smtp transport is now expanded.
3883 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3884 silently ignored. Now an error message is written to the main and panic
3887 46. There's a new ACL modifier called "logwrite" which writes to a log file
3888 as soon as it is encountered.
3890 47. Added $local_user_uid and $local_user_gid at routing time.
3892 48. Exim crashed when trying to verify a sender address that was being
3895 49. Exim was recognizing only a space character after ".include". It now also
3896 recognizes a tab character.
3898 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3899 extracting the relevant information from the specification. The man page no
3900 longer contains scrambled data for the -d option, and I've added a section
3901 at the front about calling Exim under different names.
3903 51. Added "extra_headers" argument to the "mail" command in filter files.
3905 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3908 53. Installed eximstats 1.29.
3910 54. Added transport_filter_timeout as a generic transport option.
3912 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3913 Cc: header lines. This was required by RFC 822, but it not required by RFC
3916 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3917 incoming messages that did not have them. Now it does so only if the
3918 message originates locally, that is, if there is no associated remote host
3919 address. When Resent- header lines are present, this applies to the Resent-
3920 lines rather than the non-Resent- lines.
3922 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3923 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3925 58. Messages for configuration errors now include the name of the main
3926 configuration file - useful now that there may be more than one file in a
3927 list (.included file names were always shown).
3929 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3930 for those rare installations that do not start the daemon as root or run it
3931 setuid root. I've cut out the call to initgroups() if the daemon is not
3934 60. The Exim user and group can now be bound into the binary as text strings
3935 that are looked up at the start of Exim's processing.
3937 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3939 62. Added $mailstore_basename variable.
3941 63. Installed patch to sieve.c from Michael Haardt.
3943 64. When Exim failed to open the panic log after failing to open the main log,
3944 the original message it was trying to log was written to stderr and debug
3945 output, but if they were not available (the usual case in production), it
3946 was lost. Now it is written to syslog before the two lines that record the
3947 failures to open the logs.
3949 65. Users' Exim filters run in subprocesses under the user's uid. It is
3950 possible for a "deliver" command or an alias in a "personal" command to
3951 provoke an address rewrite. If logging of address rewriting is configured,
3952 this fails because the process is not running as root or exim. There may be
3953 a better way of dealing with this, but for the moment (because 4.30 needs
3954 to be released), I have disabled address rewrite logging when running a
3955 filter in a non-root, non-exim process.
3961 1. The buildconfig auxiliary program wasn't quoting the value set for
3962 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3963 not defined. This bug was masked in 4.22 by the effect that was fixed in
3966 2. Some messages that were rejected after a message id was allocated were
3967 shown as "incomplete" by exigrep. It no longer does this for messages that
3968 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3970 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3971 have allow_domain_literals set, the ID did not get logged in the <= line.
3972 Domain literals are now always recognized in Message-ID: header lines.
3974 4. The first argument for a ${extract expansion item is the key name or field
3975 number. Leading and trailing spaces in this item were not being ignored,
3976 causing some misleading effects.
3978 5. When deliver_drop_privilege was set, single queue runner processes started
3979 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3980 same command in the process it spins off) were not dropping privilege.
3982 6. When the daemon running as "exim" started a queue runner, it always
3983 re-executed Exim in the spun-off process. This is a waste of effort when
3984 deliver_drop_privilege is set. The new process now just calls the
3985 queue-runner function directly.
3991 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3994 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3995 ignored. Though the use of -oP was forcing the writing of a pid file, it
3996 was always written to the default place.
3998 3. If the message "no IP address found for host xxxx" is generated during
3999 incoming verification, it is now followed by identification of the incoming
4000 connection (so you can more easily find what provoked it).
4002 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4004 5. Added some features to "harden" Exim a bit more against certain attacks:
4006 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4007 be put in Local/Makefile. This is like the never_users runtime option,
4008 but it cannot be overridden. The default setting is "root".
4010 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4011 prefix string with which any file named in a -C command line option
4014 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4015 is retained for -C and -D only if the caller of Exim is root. Without
4016 it, the exim user may also use -C and -D and retain privilege.
4018 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4019 command line option is disabled.
4021 6. Macro names set by the -D option must start with an upper case letter, just
4022 like macro names defined in the configuration file.
4024 7. Added "dereference=" facility to LDAP.
4026 8. Two instances of the typo "uknown" in the source files are fixed.
4028 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4029 the Configure-Makefile script screwed up while processing it.
4031 10. Incorporated PCRE 4.4.
4033 11. The SMTP synchronization check was not operating right at the start of an
4034 SMTP session. For example, it could not catch a HELO sent before the client
4035 waited for the greeting. There is now a check for outstanding input at the
4036 point when the greeting is written. Because of the duplex, asynchronous
4037 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4038 way, but not yet received, when the check is performed.
4040 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4041 on TCP/IP sockets, because this apparently causes some broken clients to
4044 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4045 unchanged) from the Cygwin maintainer.
4047 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4048 was supported instead of testing for mbx. Effectively a typo.
4050 15. The spa authenticator server code was not checking that the input it
4051 received was valid base64.
4053 16. The debug output line for the "set" modifier in ACLs was not showing the
4054 name of the variable that was being set.
4056 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4058 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4059 lookup on its own. Something else was needed to trigger the lookup. For
4060 example, a match in host_lookup or the need for a host name in a host list.
4061 Now, if $sender_host_name is referenced and the host name has not yet been
4062 looked up, a lookup is performed. If the lookup fails, the variable remains
4063 empty, and $host_lookup_failed is set to "1".
4065 19. Added "eqi" as a case-independent comparison operator.
4067 20. The saslauthd authentication condition could segfault if neither service
4068 nor realm was specified.
4070 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4071 error message that was logged was misleading, and incoming SMTP
4072 connections were dropped. The message is now more accurate, and temporary
4073 errors are given to SMTP connections.
4075 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4076 (except RSET) with a 421 error, until QUIT is received. However, it was
4077 failing to send a response to QUIT.
4079 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4080 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4081 ACL is now run after the helo_try_verify_hosts code.
4083 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4084 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4085 case-independent, so other case variants are also recognized). Apparently
4086 some systems use these upper case variants.
4088 25. If more than two messages were waiting for the same host, and a transport
4089 filter was specified for the transport, Exim sent two messages over the
4090 same TCP/IP connection, and then failed with "socket operation on non-
4091 socket" when it tried to send the third.
4093 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4095 27. The extern definition of crypt16() in expand.c was not being excluded when
4096 the OS had its own crypt16() function.
4098 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4099 as a preferred synonym for return_size_limit, both as an option and as an
4102 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4104 30. Changed the default configuration ACL to relax the local part checking rule
4105 for addresses that are not in any local domains. For these addresses,
4106 slashes and pipe symbols are allowed within local parts, but the sequence
4107 /../ is explicitly forbidden.
4109 31. SPA server authentication was not clearing the challenge buffer before
4112 32. log_message in a "warn" ACL statement was writing to the reject log as
4113 well as to the main log, which contradicts the documentation and doesn't
4114 seem right (because no rejection is happening). So I have stopped it.
4116 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4117 However, I am unable to do any testing of this.
4119 34. Fixed an infelicity in the appendfile transport. When checking directories
4120 for a mailbox, to see if any needed to be created, it was accidentally
4121 using path names with one or more superfluous leading slashes; tracing
4122 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4124 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4125 contents are added to the log line that is written for every discarded
4126 recipient. (Previously a log_message setting was ignored.)
4128 36. The ${quote: operator now quotes the string if it is empty.
4130 37. The install script runs exim in order to find its version number. If for
4131 some reason other than non-existence or emptiness, which it checks, it
4132 could not run './exim', it was installing it with an empty version number,
4133 i.e. as "exim-". This error state is now caught, and the installation is
4136 38. An argument was missing from the function that creates an error message
4137 when Exim fails to connect to the socket for saslauthd authentication.
4138 This could cause Exim to crash, or give a corrupted message.
4140 39. Added isip, isip4, and isip6 to ${if conditions.
4142 40. The ACL variables $acl_xx are now saved with the message, and can be
4143 accessed later in routers, transports, and filters.
4145 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4146 strings in the file are not string-expanded.
4148 42. If a MAIL command specified a SIZE value that was too large to fit into an
4149 int variable, the check against message_size_limit failed. Such values are
4150 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4151 day this will have to be increased, but I don't think I want to be around
4152 when emails are that large.
4159 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4160 iconv() is not standard in FreeBSD.
4162 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4163 IPv6 enabled. The observed symptom was a segmentation fault on return from
4164 the function os_common_find_running_interfaces() in src/os.c.
4166 3. In the check_special_case() function in daemon.c I had used "errno" as an
4167 argument name, which causes warnings on some systems. This was basically a
4168 typo, since it was named "eno" in the comments!
4170 4. The code that waits for the clock to tick (at a resolution of some fraction
4171 of a second) so as to ensure message-id uniqueness was always waiting for
4172 at least one whole tick, when it could have waited for less. [This is
4173 almost certainly not relevant at current processor speeds, where it is
4174 unlikely to ever wait at all. But we try to future-proof.]
4176 5. The function that sleeps for a time interval that includes fractions of a
4177 second contained a race. It did not block SIGALRM between setting the
4178 timer, and suspending (a couple of lines later). If the interval was short
4179 and the sigsuspend() was delayed until after it had expired, the suspension
4180 never ended. On busy systems this could lead to processes getting stuck for
4183 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4184 process, before it forks any delivery processes. The open lookup caching
4185 mechanism meant that the open file or database connection was passed into
4186 the delivery process. The problem was that delivery processes always tidy
4187 up cached lookup data. This could cause a problem for the next delivery
4188 process started by the queue runner, because the external queue runner
4189 process does not know about the closure. So the next delivery process
4190 still has data in the lookup cache. In the case of a file lookup, there was
4191 no problem because closing a file descriptor in a subprocess doesn't affect
4192 the parent. However, if the lookup was caching a connection to a database,
4193 the connection was closed, and the second delivery process was likely to
4194 see errors such as "PGSQL: query failed: server closed the connection
4195 unexpectedly". The problem has been fixed by closing all cached lookups
4196 in a queue runner before running a delivery process.
4198 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4199 seem to have the "const" qualifier which it has on other OS. I've
4202 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4203 *of the same type* (client or server) with the same public name that an
4204 error should be diagnosed.
4206 9. When Exim looked up a host name for an IP address, but failed to find the
4207 original IP address when looking up the host name (a safety check), it
4208 output the message "<ip address> does not match any IP for NULL", which was
4209 confusing, to say the least. The bug was that the host name should have
4210 appeared instead of "NULL".
4212 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4213 user that is built into the binary, and the -C or -D options is used, root
4214 privilege is dropped before the configuration file is read. In addition,
4215 logging is switched to stderr instead of the normal log files. If the
4216 configuration then re-defines the Exim user, the unprivileged environment
4217 is probably not what is expected, so Exim logs a panic warning message (but
4220 However, if deliver_drop_privilege is set, the unprivileged state may well
4221 be exactly what is intended, so the warning has been cut out in that case,
4222 and Exim is allowed to try to write to its normal log files.
4228 1. smtp_return_error_details was not giving details for temporary sender
4229 or receiver verification errors.
4231 2. Diagnose a configuration error if two authenticators have the same public
4234 3. Exim used not to create the message log file for a message until the first
4235 delivery attempt. This could be confusing when incoming messages were held
4236 for policy or load reasons. The message log file is now created at the time
4237 the message is received, and an initial "Received" line is written to it.
4239 4. The automatically generated man page for command line options had a minor
4240 bug that caused no ill effects; however, a more serious problem was that
4241 the procedure for building the man page automatically didn't always
4242 operate. Consequently, release 4.20 contains an out-of-date version. This
4243 shouldn't happen again.
4245 5. When building Exim with embedded Perl support, the script that builds the
4246 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4247 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4249 6. The freeze_tell option was not being used for messages that were frozen on
4250 arrival, either by an ACL or by local_scan().
4252 7. Added the smtp_incomplete_transaction log selector.
4254 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4255 was accepting AUTH without a new EHLO.
4257 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4258 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4261 10. Logging of TCP/IP connections (when configured) now happens in the main
4262 daemon process instead of the child process, so that the TCP/IP connection
4263 count is more accurate (but it can never be perfect).
4265 11. The use of "drop" in a nested ACL was not being handled correctly in the
4266 outer ACL. Now, if condition failure induced by the nested "drop" causes
4267 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4268 or "require"), the connection is dropped.
4270 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4271 that yield "discard" can now be used with an "accept" or a "discard" verb,
4272 but an error is generated for any others (because I can't see a useful way
4273 to define what should happen).
4275 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4276 are now processed in the same way as lines in the Exim configuration file.
4277 In particular, continuation lines are supported.
4279 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4281 15. Added -ti meaning -t -i.
4283 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4284 domains, and warn by logging if others are found.
4286 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4287 returned from GIFCONF when Exim is trying to find the list of interfaces on
4288 a host. The code in os.c has been modified to copy each ifreq to an aligned
4289 structure in all cases.
4291 Also, in some cases, the returned ifreq's were being copied to a 'struct
4292 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4293 means the last couple of bytes of an IPv6 address could be chopped if the
4294 ifreq contained only a normal sockaddr (14 bytes storage).
4296 18. Named domain lists were not supported in the hosts_treat_as_local option.
4297 An entry such as +xxxx was not recognized, and was treated as a literal
4300 19. Ensure that header lines added by a DATA ACL are included in the reject log
4301 if the ACL subsequently rejects the message.
4303 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4304 MD5 (which is deprecated).
4306 21. When testing a filter file using -bf, Exim was writing a message when it
4307 took the sender from a "From " line in the message, but it was not doing so
4308 when it took $return_path from a Return-Path: header line. It now does.
4310 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4311 with a valid header line field name (a series of printing characters
4312 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4314 23. Changed "disc" in the source to "disk" to conform to the documentation and
4315 the book and for uniformity.
4317 24. Ignore Sendmail's -Ooption=value command line item.
4319 25. When execve() failed while trying to run a command in a pipe transport,
4320 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4321 could be confused with a return value of 69 from the command itself. This
4322 has been changed to 127, the value the shell returns if it is asked to run
4323 a non-existent command. The wording for the related log line suggests a
4324 non-existent command as the problem.
4326 26. If received_header_text expands to an empty string, do not add a Received:
4327 header line to the message. (Well, it adds a token one on the spool, but
4328 marks it "old" so that it doesn't get used or transmitted.)
4330 27. Installed eximstats 1.28 (addition of -nt option).
4332 28. There was no check for failure on the call to getsockname() in the daemon
4333 code. This can fail if there is a shortage of resources on the system, with
4334 ENOMEM, for example. A temporary error is now given on failure.
4336 29. Contrary to the C standard, it seems that in some environments, the
4337 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4338 program. Exim now does this explicitly; it affects the formatting of
4339 timestamps using strftime().
4341 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4342 complaints. I've now initialized all the variables, to avoid this.
4344 32. Header lines added by a system filter were not being "seen" during
4345 transport-time rewrites.
4347 33. The info_callback() function passed to OpenSSL is set up with type void
4348 (*)(SSL *, int, int), as described somewhere. However, when calling the
4349 function (actually a macro) that sets it up, the type void(*)() is
4350 expected. I've put in a cast to prevent warnings from picky compilers.
4352 34. If a DNS black list lookup found a CNAME record, but there were no A
4353 records associated with the domain it pointed at, Exim crashed.
4355 35. If a DNS black list lookup returned more than one A record, Exim ignored
4356 all but the first. It now scans all returned addresses if a particular IP
4357 value is being sought. In this situation, the contents of the
4358 $dnslist_value variable are a list of all the addresses, separated by a
4361 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4362 to accept a host name and all its aliases if the forward lookup for any of
4363 them yielded the IP address of the incoming connection. Now it accepts only
4364 those names whose forward lookup yields the correct IP address. Any other
4365 names are discarded. This closes a loophole whereby a rogue DNS
4366 administrator could create reverse DNS records to break through a
4367 wildcarded host restriction in an ACL.
4369 37. If a user filter or a system filter that ran in a subprocess used any of
4370 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4371 the wrong values were passed to the pipe command ($thisaddress had the
4372 value of $0, $0 had the value of $1, etc). This bug was introduced by
4373 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4375 38. Improved the line breaking for long SMTP error messages from ACLs.
4376 Previously, if there was no break point between 40 and 75 characters, Exim
4377 left the rest of the message alone. Two changes have been made: (a) I've
4378 reduced the minimum length to 35 characters; (b) if it can't find a break
4379 point between 35 and 75 characters, it looks ahead and uses the first one
4380 that it finds. This may give the occasional overlong line, but at least the
4381 remaining text gets split now.
4383 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4384 file descriptors might be low, and that setting 1000 would always raise it.
4385 It turns out that in some environments, the limit is already over 1000 and
4386 that lowering it causes trouble. So now Exim takes care not to decrease it.
4388 40. When delivering a message, the value of $return_path is set to $sender_
4389 address at the start of routing (routers may change the value). By an
4390 oversight, this default was not being set up when an address was tested by
4391 -bt or -bv, which affected the outcome if any router or filter referred to
4394 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4395 log only when "message" or "log_message" are set. However, if one of the
4396 conditions was an address verification, or a call to a nested ACL, the
4397 messages generated by the underlying test were being passed through. This
4398 no longer happens. The underlying message is available in $acl_verify_
4399 message for both "message" and "log_message" expansions, so it can be
4400 passed through if needed.
4402 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4403 new expansion $bh_ to give the encoded byte string without charset
4404 translation. Translation happens only if iconv() is available; HAVE_ICONV
4405 indicates this at build time. HEADERS_CHARSET gives the charset to
4406 translate to; headers_charset can change it in the configuration, and
4407 "headers charset" can change it in an individual filter file.
4409 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4410 that creates RFC 2047 encoded "words" labels them as that charset instead
4411 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4412 expansion operator; (ii) when Exim creates a From: line for a local
4413 message; (iii) when a header line is rewritten to include a "phrase" part.
4415 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4416 buggy, causing it to skip the first lines of messages whose message ID
4417 ended in 'D'. This would not have bitten before Exim release 4.14, because
4418 message IDs were unlikely to end in 'D' before then. The effect was to have
4419 incorrect size information for certain domains.
4421 45. #include "config.h" was missing at the start of the crypt16.c module. This
4422 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4425 46. If there was a timeout during a "random" callout check, Exim treated it as
4426 a failure of the random address, and carried on sending RSET and the real
4427 address. If the delay was just some slowness somewhere, the response to the
4428 original RCPT would be taken as a response to RSET and so on, causing
4429 mayhem of various kinds.
4431 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4432 when I implemented it. It didn't allow for the fact that some option values
4433 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4434 the right test for positive values.
4436 48. Domain names in DNS records are case-independent. Exim always looks them up
4437 in lower case. Some resolvers return domain names in exactly the case they
4438 appear in the zone file, that is, they may contain uppercase letters. Not
4439 all resolvers do this - some return always lower case. Exim was treating a
4440 change of case by a resolver as a change of domain, similar to a widening
4441 of a domain abbreviation. This triggered its re-routing code and so it was
4442 trying to route what was effectively the same domain again. This normally
4443 caused routing to fail (because the router wouldn't handle the domain
4444 twice). Now Exim checks for this case specially, and just changes the
4445 casing of the domain that it ultimately uses when it transmits the message
4448 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4451 50. If a filter generated a file delivery with a non-absolute name (possible if
4452 no home directory exists for the router), the forbid_file option was not
4455 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4456 the existing equality matching.
4458 52. Exim was using ints instead of ino_t variables in some places where it was
4459 dealing with inode numbers.
4461 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4462 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4463 called TMPDIR, and if it finds it is different, it changes its value.
4465 54. The smtp_printf() function is now made available to local_scan() so
4466 additional output lines can be written before returning. There is also an
4467 smtp_fflush() function to enable the detection of a dropped connection.
4468 The variables smtp_input and smtp_batched_input are exported to
4471 55. Changed the default runtime configuration: the message "Unknown user"
4472 has been removed from the ACL, and instead placed on the localuser router,
4473 using the cannot_route_message feature. This means that any verification
4474 failures that generate their own messages won't get overridden. Similarly,
4475 the "Unrouteable address" message that was in the ACL for unverifiable
4476 relay addresses has also been removed.
4478 56. Added hosts_avoid_esmtp to the smtp transport.
4480 57. The exicyclog script was not checking for the esoteric option
4481 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4482 will work only if exicyclog is run under the appropriate euid.
4484 58. Following a discussion on the list, the rules by which Exim recognises line
4485 endings on incoming messages have been changed. The -dropcr and drop_cr
4486 options are now no-ops, retained only for backwards compatibility. The
4487 following line terminators are recognized: LF CRLF CR. However, special
4488 processing applies to CR:
4490 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4491 nor a local message in the state where . is a terminator.
4493 (ii) If a bare CR is encountered in a header line, an extra space is added
4494 after the line terminator so as not to end the header. The reasoning
4495 behind this is that bare CRs in header lines are most likely either
4496 to be mistakes, or people trying to play silly games.
4498 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4499 was being incorrectly given as 18 bytes larger than it should have been.
4500 This is a VOB (very old bug).
4502 60. This may never have affected anything current, but just in case it has:
4503 When the local host is found other than at the start of a list of hosts,
4504 the local host, those with the same MX, and any that follow, are discarded.
4505 When the list in question was part of a longer list of hosts, the following
4506 hosts (not currently being processed) were also being discarded. This no
4507 longer happens. I'm not sure if this situation could ever has previously
4510 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4513 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4514 header. This is something that is recommended in a new Internet Draft, and
4515 is something that is documented as being done by Sendmail. There are two
4516 possible values. For messages generated by the autoreply transport, Exim
4519 Auto-Submitted: auto-replied
4521 whereas for all other generated messages (e.g. bounces) it adds
4523 Auto-Submitted: auto-generated
4525 63. The "personal" condition in filters now includes a test for the
4526 Auto-Submitted: header. If it contains the string "auto-" the message it
4527 not considered personal.
4529 64. Added rcpt_include_affixes as a generic transport option.
4531 65. Added queue_only_override (default true).
4533 66. Added the syslog_duplication option.
4535 67. If what should have been the first header line of a message consisted of
4536 a space followed by a colon, Exim was mis-interpreting it as a header line.
4537 It isn't of course - it is syntactically invalid and should therefore be
4538 treated as the start of the message body. The misbehaviour could have
4539 caused a number of strange effects, including loss of data in subsequent
4540 header lines, and spool format errors.
4542 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4543 client host had authenticated. This control can now be exercised by an ACL
4544 for more flexibility.
4546 69. By default, callouts do not happen when testing with -bh. There is now a
4547 variant, -bhc, which does actually run the callout code, including
4548 consulting and updating the callout cache.
4550 70. Added support for saslauthd authentication, courtesy of Alexander
4553 71. If statvfs() failed on the spool or log directories while checking their
4554 size for availability, Exim confusingly gave the error "space shortage".
4555 Furthermore, in debugging mode it crashed with a floating point exception.
4556 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4557 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4558 serious problem, Exim now writes to the main and panic logs when this
4559 happens, with details of the failure. It then refuses to accept the
4560 incoming message, giving the message "spool directory problem" or "log
4561 directory problem" with a 421 code for SMTP messages.
4563 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4564 0, 1, and 2 exist, because some OS complain for execs without them (see
4565 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4566 descriptors. However, the code omitted to check that the open succeeded,
4567 causing mysterious errors if for some reason the permissions on /dev/null
4568 got screwed. Now Exim writes a message to the main and panic logs, and
4569 bombs out if it can't open /dev/null.
4571 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4572 interact so that it is all more flexible. It is supposed to remain
4573 backwards compatible. Also added extra_local_interfaces.
4575 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4576 to bomb out with an assertion failure - to the client this appears as a
4577 connection drop. This problem occurs in the part of the code that was taken
4578 from the Samba project. Fortunately, the assertion is in a very simple
4579 function, so I have fixed this by reproducing the function inline in the
4580 one place where it is called, and arranging for authentication to fail
4581 instead of killing the process with assert().
4583 75. The SPA client code was not working when the server requested OEM rather
4584 than Unicode encoding.
4586 76. Added code to make require_files with a specific uid setting more usable in
4587 the case where statting the file as root fails - usually a non-root-mounted
4588 NFS file system. When this happens and the failure is EACCES, Exim now
4589 forks a subprocess and does the per-uid checking as the relevant uid.
4591 77. Added process_log_path.
4593 78. If log_file_path was not explicitly set, a setting of check_log_space or
4594 check_log_inodes was ignored.
4596 79. If a space check for the spool or log partitions fails, the incident is now
4597 logged. Of course, in the latter case the data may get lost...
4599 80. Added the %p formatting code to string_format() so that it can be used to
4600 print addresses in debug_print(). Adjusted all the address printing in the
4601 debugging in store.c to use %p rather than %d.
4603 81. There was a concern that a line of code in smtp_in.c could overflow a
4604 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4605 initially expressed, the concern was not well-founded, because trailing
4606 spaces are removed early. However, if the trailing spaces were followed by
4607 a NULL, they did not get removed, so the overflow was possible. Two fixes
4610 (a) I re-wrote the offending code in a cleaner fashion.
4611 (b) If an incoming SMTP command contains a NULL character, it is rejected
4614 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4615 runs initgroups(), so that if the Exim user is in any additional groups,
4616 they will be used during message reception.
4622 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob