1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.479 2007/02/14 14:59:01 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
63 PH/11 Added control=no_pipelining.
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
72 PH/14 Added log_selector = +pid.
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
76 PH/16 Add ${if forany and ${if forall.
78 PH/17 Added dsn_from option to vary the From: line in DSNs.
80 PH/18 Flush SMTP output before performing a callout, unless control =
81 no_callout_flush is set.
83 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
84 was true (the default) a successful delivery failed to delete the retry
85 item, thus causing premature timeout of the address. The bug is now
88 PH/20 Added hosts_avoid_pipelining to the smtp transport.
90 PH/21 Long custom messages for fakedefer and fakereject are now split up
91 into multiline reponses in the same way that messages for "deny" and
92 other ACL rejections are.
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 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
105 ${reduce, with only minor "tidies".
107 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
113 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
114 fixed by 4.65/MH/01 (is this a record?) are fixed:
116 (i) An empty string was always treated as zero by the numeric comparison
117 operators. This behaviour has been restored.
119 (ii) It is documented that the numeric comparison operators always treat
120 their arguments as decimal numbers. This was broken in that numbers
121 starting with 0 were being interpreted as octal.
123 While fixing these problems I realized that there was another issue that
124 hadn't been noticed. Values of message_size_limit (both the global option
125 and the transport option) were treated as octal if they started with 0.
126 The documentation was vague. These values are now always treated as
127 decimal, and I will make that clear in the documentation.
133 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
134 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
137 MH/01 Don't check that the operands of numeric comparison operators are
138 integers when their expansion is in "skipping" mode (fixes bug
139 introduced by 4.64-PH/07).
141 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
142 child addresses, Exim now panics and dies. Previously, because the count
143 is held in a short int, deliveries were likely to be lost. As such a
144 large number of recipients for a single message is ridiculous
145 (performance will be very, very poor), I have chosen to impose a limit
146 rather than extend the field.
152 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
153 leftover -K file (the existence of which was triggered by #402).
154 While we were at it, introduced process PID as part of the -K
155 filename. This should rule out race conditions when creating
158 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
159 processing considerably. Previous code took too long for large mails,
160 triggering a timeout which in turn triggers #401.
162 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
163 in the DK code in transports.c. sendfile() is not really portable,
164 hence the _LINUX specificness.
166 TF/01 In the add_headers option to the mail command in an Exim filter,
167 there was a bug that Exim would claim a syntax error in any
168 header after the first one which had an odd number of characters
171 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
172 callout verification, Exim cached a "reject" for the entire domain. This
173 is correct for most verifications, but it is not correct for a recipient
174 verification with use_sender or use_postmaster set, because in that case
175 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
176 case of MAIL FROM:<> rejection from other early rejections (e.g.
177 rejection of HELO). When verifying a recipient using a non-null MAIL
178 address, the cache is ignored if it shows MAIL FROM:<> rejection.
179 Whatever the result of the callout, the value of the domain cache is
180 left unchanged (for any other kind of callout, getting as far as trying
181 RCPT means that the domain itself is ok).
183 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
186 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
187 manifest itself as EPIPE rather than ECONNECT. When tidying away a
188 session, the daemon ignores ECONNECT errors and logs others; it now
189 ignores EPIPE as well.
191 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
192 (quoted-printable decoding).
194 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
195 later the small subsequent patch to fix an introduced bug.
197 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
199 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
201 PH/08 An error is now given if message_size_limit is specified negative.
203 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
204 to be given (somewhat) arbitrary names.
206 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
209 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
210 miscellaneous code fixes
212 PH/10 Added the log_reject_target ACL modifier to specify where to log
215 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
216 hostname. This is wrong, because it relates to the incoming message (and
217 probably the interface on which it is arriving) and not to the outgoing
218 callout (which could be using a different interface). This has been
219 changed to use the value of the helo_data option from the smtp transport
220 instead - this is what is used when a message is actually being sent. If
221 there is no remote transport (possible with a router that sets up host
222 addresses), $smtp_active_hostname is used.
224 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
225 tweaks were necessary in order to get it to work (see also 21 below):
226 (a) The code assumed that strncpy() returns a negative number on buffer
227 overflow, which isn't the case. Replaced with Exim's string_format()
229 (b) There were several signed/unsigned issues. I just did the minimum
230 hacking in of casts. There is scope for a larger refactoring.
231 (c) The code used strcasecmp() which is not a standard C function.
232 Replaced with Exim's strcmpic() function.
233 (d) The code set only $1; it now sets $auth1 as well.
234 (e) A simple test gave the error "authentication client didn't specify
235 service in request". It would seem that Dovecot has changed its
236 interface. Fortunately there's a specification; I followed it and
237 changed what the client sends and it appears to be working now.
239 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
242 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
243 address A is aliased to B and C, where B exists and C does not. Without
244 -v the output is "A verified" because verification stops after a
245 successful redirection if more than one address is generated. However,
246 with -v the child addresses are also verified. Exim was outputting "A
247 failed to verify" and then showing the successful verification for C,
248 with its parentage. It now outputs "B failed to verify", showing B's
249 parentage before showing the successful verification of C.
251 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
252 look up a TXT record in a specific list after matching in a combined
255 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
256 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
257 they consult the DNS. I had assumed they would set it the way they
258 wanted; and indeed my experiments on Linux seem to show that in some
259 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
260 To be on the safe side, however, I have now made the interface to
261 host_find_byname() similar to host_find_bydns(), with an argument
262 containing the DNS resolver options. The host_find_byname() function now
263 sets these options at its start, just as host_find_bydns() does. The smtp
264 transport options dns_qualify_single and dns_search_parents are passed to
265 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
266 of host_find_byname() use the default settings of RES_DEFNAMES
267 (qualify_single) but not RES_DNSRCH (search_parents).
269 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
270 spool_read_header() do less string testing, by means of a preliminary
271 switch on the second character of optional "-foo" lines. (This is
272 overdue, caused by the large number of possibilities that now exist.
273 Originally there were few.) While I was there, I also converted the
274 str(n)cmp tests so they don't re-test the leading "-" and the first
275 character, in the hope this might squeeze out yet more improvement.
277 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
278 flag allowing group syntax was set by the header_syntax check but not
279 turned off, possible causing trouble later; (2) The flag was not being
280 set at all for the header_verify test, causing "group"-style headers to
281 be rejected. I have now set it in this case, and also caused header_
282 verify to ignore an empty address taken from a group. While doing this, I
283 came across some other cases where the code for allowing group syntax
284 while scanning a header line wasn't quite right (mostly, not resetting
285 the flag correctly in the right place). These bugs could have caused
286 trouble for malformed header lines. I hope it is now all correct.
288 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
289 with the "reply" argument non-NULL. The code, however (which originally
290 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
291 but it didn't always do it. This confused somebody who was copying the
292 code for some other use. I have removed all the tests.
294 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
295 feature that was used to support insecure browsers during the U.S. crypto
296 embargo. It requires special client support, and Exim is probably the
297 only MTA that supported it -- and would never use it because real RSA is
298 always available. This code has been removed, because it had the bad
299 effect of slowing Exim down by computing (never used) parameters for the
300 RSA_EXPORT functionality.
302 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
303 authenticator to fail if there's a tab character in the incoming data
304 (there should never be unless someone is messing about, as it's supposed
305 to be base64-encoded). Also added, on Timo's advice, the "secured" option
306 if the connection is using TLS or if the remote IP is the same as the
307 local IP, and the "valid-client-cert option" if a client certificate has
310 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
311 authenticators. This can be used for authorization after authentication
312 succeeds. (In the case of plaintext, it servers for both authentication
315 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
316 if any retry times were supplied.
318 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
319 connection, where there is no client IP address to check. In this
320 situation, the verify now always succeeds.
322 PH/25 Applied John Jetmore's -Mset patch.
324 PH/26 Added -bem to be like -Mset, but loading a message from a file.
326 PH/27 In a string expansion for a processed (not raw) header when multiple
327 headers of the same name were present, leading whitespace was being
328 removed from all of them, but trailing whitespace was being removed only
329 from the last one. Now trailing whitespace is removed from each header
330 before concatenation. Completely empty headers in a concatenation (as
333 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
334 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
336 PH/29 [Removed. This was a change that I later backed out, and forgot to
337 correct the ChangeLog entry (that I had efficiently created) before
338 committing the later change.]
340 PH/30 Exim was sometimes attempting to deliver messages that had suffered
341 address errors (4xx response to RCPT) over the same connection as other
342 messages routed to the same hosts. Such deliveries are always "forced",
343 so retry times are not inspected. This resulted in far too many retries
344 for the affected addresses. The effect occurred only when there were more
345 hosts than the hosts_max_try setting in the smtp transport when it had
346 the 4xx errors. Those hosts that it had tried were not added to the list
347 of hosts for which the message was waiting, so if all were tried, there
348 was no problem. Two fixes have been applied:
350 (i) If there are any address or message errors in an SMTP delivery, none
351 of the hosts (tried or untried) are now added to the list of hosts
352 for which the message is waiting, so the message should not be a
353 candidate for sending over the same connection that was used for a
354 successful delivery of some other message. This seems entirely
355 reasonable: after all the message is NOT "waiting for some host".
356 This is so "obvious" that I'm not sure why it wasn't done
357 previously. Hope I haven't missed anything, but it can't do any
358 harm, as the worst effect is to miss an optimization.
360 (ii) If, despite (i), such a delivery is accidentally attempted, the
361 routing retry time is respected, so at least it doesn't keep
362 hammering the server.
364 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
365 in ${readsocket because some servers need this prod.
367 PH/32 Added some extra debug output when updating a wait-xxx database.
369 PH/33 The hint "could be header name not terminated by colon", which has been
370 given for certain expansion errors for a long time, was not being given
371 for the ${if def:h_colon_omitted{... case.
373 PH/34 The spec says: "With one important exception, whenever a domain list is
374 being scanned, $domain contains the subject domain." There was at least
375 one case where this was not true.
377 PH/35 The error "getsockname() failed: connection reset by peer" was being
378 written to the panic log as well as the main log, but it isn't really
379 panic-worthy as it just means the connection died rather early on. I have
380 removed the panic log writing for the ECONNRESET error when getsockname()
383 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
384 runs only) independently of the message's sender address. This meant
385 that, if the 4xx error was in fact related to the sender, a different
386 message to the same recipient with a different sender could confuse
387 things. In particualar, this can happen when sending to a greylisting
388 server, but other circumstances could also provoke similar problems.
389 I have changed the default so that the retry time for these errors is now
390 based a combination of the sender and recipient addresses. This change
391 can be overridden by setting address_retry_include_sender=false in the
394 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
395 remote server are returned as part of bounce messages. This was not
396 happening for LMTP over a pipe (the lmtp transport), but now it is the
397 same for both kinds of LMTP.
399 PH/38 Despite being documented as not happening, Exim was rewriting addresses
400 in header lines that were in fact CNAMEs. This is no longer the case.
402 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
403 and queue runs started by the daemon processed all messages. This has
404 been fixed so that -R and -S can now usefully be given with -q<time>.
406 PH/40 Import PCRE release 6.7 (fixes some bugs).
408 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
410 PH/42 Give an error if -q is specified more than once.
412 PH/43 Renamed the variables $interface_address and $interface_port as
413 $received_ip_address and $received_port, to make it clear that these
414 values apply to message reception, and not to the outgoing interface when
415 a message is delivered. (The old names remain recognized, of course.)
417 PH/44 There was no timeout on the connect() call when using a Unix domain
418 socket in the ${readsocket expansion. There now is.
420 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
421 be meaningful with "accept".
423 SC/01 Eximstats V1.43
424 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
426 SC/02 Eximstats V1.44
427 Use a glob alias rather than an array ref in the generated
428 parser. This improves both readability and performance.
430 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
431 Collect SpamAssassin and rejection statistics.
432 Don't display local sender or destination tables unless
433 there is data to show.
434 Added average volumes into the top table text output.
436 SC/04 Eximstats V1.46
437 Collect data on the number of addresses (recipients)
438 as well as the number of messages.
440 SC/05 Eximstats V1.47
441 Added 'Message too big' to the list of mail rejection
442 reasons (thanks to Marco Gaiarin).
444 SC/06 Eximstats V1.48
445 Mainlog lines which have GMT offsets and are too short to
446 have a flag are now skipped.
448 SC/07 Eximstats V1.49 (Alain Williams)
449 Added the -emptyok flag.
451 SC/08 Eximstats V1.50
452 Fixes for obtaining the IP address from reject messages.
454 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
455 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
456 whitesspace changes from 4.64-PH/27
458 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
461 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
462 are found, allow negative numbers in numeric criteria)
464 JJ/06 exipick.20061117.2, added new $message_body_missing variable
466 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
467 to match changes made in 4.64-PH/43
469 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
471 PH/47 Put in an explicit test for a DNS lookup of an address record where the
472 "domain" is actually an IP address, and force a failure. This locks out
473 those revolvers/nameservers that support "A-for-A" lookups, in
474 contravention of the specifications.
476 PH/48 When a host name was looked up from an IP address, and the subsequent
477 forward lookup of the name timed out, the host name was left in
478 $sender_host_name, contrary to the specification.
480 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
481 restricted to single-key lookups, Exim was not diagnosing an error if
482 * or *@ was used with a query-style lookup.
484 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
486 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
487 long ago, but noone interested enough thought of it. Let's just say that
488 the "1.1" means that there are some new functions that weren't there at
489 some point in the past.
491 PH/51 Error processing for expansion failure of helo_data from an smtp
492 transport during callout processing was broken.
494 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
495 tested/used via the -bh/-bhc/-bs options.
497 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
498 bug, fixed in subsequent PCRE releases).
500 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
501 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
503 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
509 SC/01 Use a glob alias rather than an array ref in eximstats generated
510 parser. This improves both readability and performance.
512 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
513 Don't display local sender or destination tables in eximstats unless
514 there is data to show.
515 Added average volumes into the eximstats top table text output.
517 SC/03 Collect data on the number of addresses (recipients) as well
518 as the number of messages in eximstats.
520 TF/01 Correct an error in the documentation for the redirect router. Exim
521 does (usually) call initgroups() when daemonizing.
523 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
524 with consistent privilege compared to when running as a daemon.
526 TF/03 Note in the spec that $authenticated_id is not set for local
527 submissions from trusted users.
529 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
530 Thanks to Dean Brooks <dean@iglou.com> for the patch.
532 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
533 by adding some example configuration directives to the default
534 configuration file. A little bit of work is required to uncomment the
535 directives and define how usernames and passwords are checked, but
536 there is now a framework to start from.
538 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
539 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
540 without this. I don't know how relevant this is to other LDAP libraries.
542 PH/02 Add the verb name to the "unknown ACL verb" error.
544 PH/03 Magnus Holmgren's patch for filter_prepend_home.
546 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
548 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
549 directory not expanded when it should be if an expanded home directory
550 was set for the address (which is overridden by the transport).
552 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
555 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
556 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
557 because it is too late at that time, and has no effect.
559 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
560 security issue with \' (bugzilla #107). I could not use the
561 PQescapeStringConn() function, because it needs a PGconn value as one of
564 PH/08 When testing addresses using -bt, indicate those final addresses that
565 are duplicates that would not cause an additional delivery. At least one
566 person was confused, thinking that -bt output corresponded to deliveries.
567 (Suppressing duplicates isn't a good idea as you lose the information
568 about possibly different redirections that led to the duplicates.)
570 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
571 systems where poll() doesn't work, in particular OS X.
573 PH/10 Added more information to debugging output for retry time not reached.
575 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
576 operations in malware.c.
578 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
581 PH/13 If write_rejectlog was set false when logging was sent to syslog with
582 syslog_duplication set false, log lines that would normally be written
583 both the the main log and to the reject log were not written to syslog at
586 PH/14 In the default configuration, change the use of "message" in ACL warn
587 statements to "add_header".
589 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
590 not followed by a command (e.g. "seen endif").
592 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
593 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
596 PH/17 Added extra conditions to the default value of delay_warning_condition
600 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
601 { match{$h_precedence:}{(?i)bulk|list|junk} } \
602 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
605 The Auto-Submitted: and various List- headers are standardised, whereas I
606 don't think Precedence: ever was.
608 PH/18 Refactored debugging code in route_finduser() to show more information,
609 in particular, the error code if getpwnam() issues one.
611 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
612 This is apparently needed in addition to the PH/07 change above to avoid
613 any possible encoding problems.
615 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
616 but not after initializing Perl.
618 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
619 output them only if debugging. By default they are written stderr,
620 apparently, which is not desirable.
622 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
625 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
628 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
630 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
631 authenticated or an ident call has been made. Suppress the default
632 values for $authenticated_id and $authenticated_sender (but permit -oMai
633 and -oMas) when testing with -bh.
635 PH/24 Re-jigged the order of the tests in the default configuration so that the
636 tests for valid domains and recipients precede the DNS black list and CSA
637 tests, on the grounds that those ones are more expensive.
639 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
640 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
641 This bug exists in every version of Exim that I still have, right back to
644 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
645 However, an attempt to turn on case-sensitivity in a regex key by
646 including (?-i) didn't work because the subject string was already
647 lowercased, and the effects were non-intuitive. It turns out that a
648 one-line patch can be used to allow (?-i) to work as expected.
654 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
655 other effects) broke the use of negated acl sub-conditions.
657 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
660 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
661 "Deny" causes Exim to reject the incoming connection with a 554 error.
662 Unfortunately, if there is a major crisis, such as a disk failure,
663 tcp-wrappers gives "deny", whereas what one would like would be some
664 kind of temporary error. A kludge has been added to help with this.
665 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
666 554 error is used if errno is still zero or contains ENOENT (which occurs
667 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
670 PH/03 Add -lutil to the default FreeBSD LIBS setting.
672 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
673 errors. Otherwise a message that provokes a temporary error (when other
674 messages do not) can cause a whole host to time out.
676 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
677 addresses were routed directly to files or pipes from a redirect router.
678 File deliveries just didn't batch; pipe deliveries might have suffered
681 PH/06 A failure to get a lock for a hints database would erroneously always say
682 "Failed to get write lock", even when it was really a read lock.
684 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
685 of 0600. This has been changed to use the value of the lockfile_mode
686 option (which defaults to 0600).
688 PH/08 Applied small patch from the Sieve maintainer.
690 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
691 folder from quota calculations, a direct delivery into this folder messed
692 up the contents of the maildirsize file. This was because the regex was
693 used only to exclude .Trash (or whatever) when the size of the mailbox
694 was calculated. There was no check that a delivery was happening into an
695 excluded directory. This bug has been fixed by ignoring all quota
696 processing for deliveries into excluded directories.
698 PH/10 Added the maildirfolder_create_regex option to appendfile.
704 PH/01 The code for finding all the local interface addresses on a FreeBSD
705 system running IPv6 was broken. This may well have applied to all BSD
706 systems, as well as to others that have similar system calls. The broken
707 code found IPv4 interfaces correctly, but gave incorrect values for the
708 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
709 that it would not match correctly against @[] and not recognize the IPv6
712 PH/02 The ipliteral router was not recognizing addresses of the form user@
713 [ipv6:....] because it didn't know about the "ipv6:" prefix.
715 PH/03 Added disable_ipv6.
717 PH/04 Changed $reply_address to use the raw form of the headers instead of the
718 decoded form, because it is most often used to construct To: headers
719 lines in autoreplies, and the decoded form may well be syntactically
720 invalid. However, $reply_address has leading white space removed, and all
721 newlines turned into spaces so that the autoreply transport does not
724 PH/05 If group was specified without a user on a router, and no group or user
725 was specified on a transport, the group from the router was ignored.
727 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
728 for visible compile-time settings that can be used to change these
729 numbers, for those that want even more. Backwards compatibility with old
730 spool files has been maintained. However, going back to a previous Exim
731 release will lost any variables that are in spool files.
733 PH/07 Two small changes when running in the test harness: increase delay when
734 passing a TCP/IP connection to a new process, in case the original
735 process has to generate a bounce, and remove special handling of
736 127.0.0.2 (sic), which is no longer necessary.
738 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
739 be the same on different OS.
741 PH/09 Moved a debug statement in filter processing to avoid a race problem when
744 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
745 whether --show-vars was specified or not
747 JJ/02 exipick: Added support for new ACL variable spool format introduced
750 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
751 syntactically invalid From: or Reply-to: line, and a filter used this to
752 generate an autoreply, and therefore failed to obtain an address for the
753 autoreply, Exim could try to deliver to a non-existent relative file
754 name, causing unrelated and misleading errors. What now happens is that
755 it logs this as a hard delivery error, but does not attempt to create a
758 PH/11 The exinext utility has a -C option for testing purposes, but although
759 the given file was scanned by exinext itself; it wasn't being passed on
760 when Exim was called.
762 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
763 an end-of-file indication when reading a command response.
765 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
766 compiled. In many other places in Exim, IPv6 addresses are always
767 recognized, so I have changed this. It also means that IPv4 domain
768 literals of the form [IPV4:n.n.n.n] are now always recognized.
770 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
771 used if the router is not running as root, for example, when verifying at
772 ACL time, or when using -bh. The debugging output from this situation was
773 non-existent - all you got was a failure to exec. I have made two
776 (a) Failures to set uid/gid, the current directory, or a process leader
777 in a subprocess such as that created by queryprogram now generate
778 suitable debugging ouput when -d is set.
780 (b) The queryprogram router detects when it is not running as root,
781 outputs suitable debugging information if -d is set, and then runs
782 the subprocess without attempting to change uid/gid.
784 PH/15 Minor change to Makefile for building test_host (undocumented testing
787 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
788 additional section of a DNS packet that returns MX or SRV records.
789 Instead, it always explicitly searches for A/AAAA records. This avoids
790 major problems that occur when a DNS server includes only records of one
791 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
792 fixed another bug: if SRV records were looked up and the corresponding
793 address records were *not* found in the additional section, the port
794 values from the SRV records were lost.
796 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
797 using the correct key (the original address) when searching the retry
798 rules in order to find which one to use for generating the retry hint.
800 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
801 adding the default one. Similarly, if it contains a Reply-To: header, the
802 errors_reply_to option, if set, is not used.
804 PH/19 When calculating a retry time, Exim used to measure the "time since
805 failure" by looking at the "first failed" field in the retry record. Now
806 it does not use this if it is later than than the arrival time of the
807 message. Instead it uses the arrival time. This makes for better
808 behaviour in cases where some deliveries succeed, thus re-setting the
809 "first failed" field. An example is a quota failure for a huge message
810 when small messages continue to be delivered. Without this change, the
811 "time since failure" will always be short, possible causing more frequent
812 delivery attempts for the huge message than are intended.
813 [Note: This change was subsequently modified - see PH/04 for 4.62.]
815 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
816 $1, $2, $3) because the numerical variables can be reset during some
817 expansion items (e.g. "match"), thereby losing the authentication data.
819 PH/21 Make -bV show the size of off_t variables so that the test suite can
820 decide whether to run tests for quotas > 2G.
822 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
823 mailbox_size, and mailbox_filecount in the appendfile transport. If a
824 filecount value is greater than 2G or if a quota value is greater than 2G
825 on a system where the size of off_t is not greater than 4, a panic error
828 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
829 never match. The debug and -bh output now contains an explicit error
830 message indicating a malformed IPv4 address or mask.
832 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
833 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
836 PH/25 Do not write to syslog when running in the test harness. The only
837 occasion when this arises is a failure to open the main or panic logs
838 (for which there is an explicit test).
840 PH/26 Added the /no_tell option to "control=freeze".
842 PH/27 If a host name lookup failed very early in a connection, for example, if
843 the IP address matched host_lookup and the reverse lookup yielded a name
844 that did not have a forward lookup, an error message of the form "no IP
845 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
846 could be logged. Now it outputs the IP address instead of "NULL".
848 PH/28 An enabling patch from MH: add new function child_open_exim2() which
849 allows the sender and the authenticated sender to be set when
850 submitting a message from within Exim. Since child_open_exim() is
851 documented for local_scan(), the new function should be too.
853 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
854 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
855 results in an empty string is now treated as unset.
857 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
859 PH/31 Added sender_verify_fail logging option.
861 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
862 needed by RFC 822 but not by RFC 2822 was commented out. I have now
863 tidied the source and removed it altogether.
865 PH/33 When a queue run was abandoned because the load average was too high, a
866 log line was always written; now it is written only if the queue_run log
867 selector is set. In addition, the log line for abandonment now contains
868 information about the queue run such as the pid. This is always present
869 in "start" and "stop" lines but was omitted from the "abandon" line.
871 PH/34 Omit spaces between a header name and the colon in the error message that
872 is given when verify = headers_syntax fails (if there are lots of them,
873 the message gets confusing).
875 PH/35 Change the default for dns_check_names_pattern to allow slashes within
876 names, as there are now some PTR records that contain slashes. This check
877 is only to protect against broken name servers that fall over on strange
878 characters, so the fact that it applies to all lookups doesn't matter.
880 PH/36 Now that the new test suite is complete, we can remove some of the
881 special code in Exim that was needed for the old test suite. For example,
882 sorting DNS records because real resolvers return them in an arbitrary
883 order. The new test suite's fake resolver always returns records in the
886 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
887 bounces) except when queue_only is set, to avoid logging races between
888 the different processes.
890 PH/38 Panic-die if .include specifies a non-absolute path.
892 PH/39 A tweak to the "H" retry rule from its user.
894 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
895 a label. They prevented compilation on older perls.
897 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
898 a warning to be raised on newish perls.
900 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
901 on queue. Changes to match documented behaviour of showing count of
902 messages matching specified criteria.
904 PH/40 Changed the default ident timeout from 30s to 5s.
906 PH/41 Added support for the use of login_cap features, on those BSD systems
907 that have them, for controlling the resources used by pipe deliveries.
909 PH/42 The content-scanning code uses fopen() to create files in which to put
910 message data. Previously it was not paying any attention to the mode of
911 the files. Exim runs with umask(0) because the rest of the code creates
912 files with open(), and sets the required mode explicitly. Thus, these
913 files were ending up world-writeable. This was not a big issue, because,
914 being within the spool directory, they were not world-accessible. I have
915 created a function called modefopen, which takes an additional mode
916 argument. It sets umask(777), creates the file, chmods it to the required
917 mode, then resets the umask. All the relevant calls to fopen() in the
918 content scanning code have been changed to use this function.
920 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
921 to 24 hours. This avoids potential overflow problems when processing G
922 and H retry rules. I suspect nobody ever tinkers with this value.
924 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
926 PH/45 When the plaintext authenticator is running as a client, the server's
927 challenges are checked to ensure they are valid base64 strings. By
928 default, the authentication attempt is cancelled if an invalid string is
929 received. Setting client_ignore_invalid_base64 true ignores these errors.
930 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
931 they are received. Thus, the responses can be made to depend on the
932 challenges. If an invalid string is ignored, an empty string is placed in
935 PH/46 Messages that are created by the autoreply transport now contains a
936 References: header, in accordance with RFCs 2822 and 3834.
938 PH/47 Added authenticated_sender_force to the smtp transport.
940 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
942 PH/49 Installed latest patch from the Sieve maintainer.
944 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
945 was also set, the appendfile transport was unnecessarily scanning a
946 directory of message files (e.g. for maildir delivery) to find the count
947 of files (along with the size), even though it did not need this
948 information. It now does the scan only if it needs to find either the
949 size of the count of files.
951 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
953 PH/52 Two bugs concerned with error handling when the smtp transport is
956 (i) Exim was not creating retry information for temporary errors given
957 for individual recipients after the DATA command when the smtp transport
958 was used in LMTP mode. This meant that they could be retried too
959 frequently, and not timed out correctly.
961 (ii) Exim was setting the flag that allows error details to be returned
962 for LMTP errors on RCPT commands, but not for LMTP errors for individual
963 recipients that were returned after the DATA command.
965 PH/53 This is related to PH/52, but is more general: for any failing address,
966 when detailed error information was permitted to be returned to the
967 sender, but the error was temporary, then after the final timeout, only
968 "retry timeout exceeded" was returned. Now it returns the full error as
969 well as "retry timeout exceeded".
971 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
972 do this, and (what is worse) MTAs that accept it.
974 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
975 will now be deprecated.
977 PH/56 New os.c-cygwin from the Cygwin maintainer.
979 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
980 formats (previously only available in exim formats via -bpr, -bpru,
981 and -bpra. Now also available in native and exiqgrep formats)
983 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
984 with very large, slow to parse queues
986 JJ/08 exipick: added ! as generic prefix to negate any criteria format
988 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
990 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
991 responses to authentication challenges, though it was showing the
992 challenges; (ii) I've removed the CR characters from the debug output for
995 PH/58 Allow for the insertion of a newline as well as a space when a string
996 is turned into more than one encoded-word during RFC 2047 encoding. The
997 Sieve code now uses this.
999 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1000 data_4xx, lost_connection, tls_required.
1002 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1003 message was being sent as an SMTP response.
1005 PH/61 Add -l and -k options to exicyclog.
1007 PH/62 When verifying, if an address was redirected to one new address, so that
1008 verification continued, and the new address failed or deferred after
1009 having set something in $address_data, the value of $address_data was not
1010 passed back to the ACL. This was different to the case when no
1011 redirection occurred. The value is now passed back in both cases.
1013 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1014 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1015 use login_cap.h, so on its own it isn't the distinguishing feature. The
1016 new name refers directly to the setclassresources() function.
1018 PH/65 Added configuration files for NetBSD3.
1020 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1022 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1023 is preferred over IPv4.
1025 PH/68 The bounce_return_message and bounce_return_body options were not being
1026 honoured for bounces generated during the reception of non-SMTP messages.
1027 In particular, this applied to messages rejected by the ACL. This bug has
1028 been fixed. However, if bounce_return_message is true and bounce_return_
1029 body is false, the headers that are returned for a non-SMTP message
1030 include only those that have been read before the error was detected.
1031 (In the case of an ACL rejection, they have all been read.)
1033 PH/69 The HTML version of the specification is now built in a directory called
1034 spec_html instead of spec.html, because the latter looks like a path with
1035 a MIME-type, and this confuses some software.
1037 PH/70 Catch two compiler warnings in sieve.c.
1039 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1040 function verify_get_ident() calls ip_connect() to connect a socket, but
1041 if the "connect()" function timed out, ip_connect() used to close the
1042 socket. However, verify_get_ident() also closes the socket later, and in
1043 between Exim writes to the log, which may get opened at this point. When
1044 the socket was closed in ip_connect(), the log could get the same file
1045 descriptor number as the socket. This naturally causes chaos. The fix is
1046 not to close the socket in ip_connect(); the socket should be closed by
1047 the function that creates it. There was only one place in the code where
1048 this was missing, in the iplookup router, which I don't think anybody now
1049 uses, but I've fixed it anyway.
1051 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1052 well as to direct DNS lookups. Otherwise the handling of names in host
1053 lists is inconsistent and therefore confusing.
1059 PH/01 Two changes to the default runtime configuration:
1061 (1) Move the checks for relay_from_hosts and authenticated clients from
1062 after to before the (commented out) DNS black list checks.
1064 (2) Add control=submission to the relay_from_hosts and authenticated
1065 clients checks, on the grounds that messages accepted by these
1066 statements are most likely to be submissions.
1068 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1070 (1) Generate an error if the third argument for the ${prvs expansion is
1073 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1076 (3) Reset the variables that are obtained from the first argument of
1077 ${prvscheck and used in the second argument before leaving the code,
1078 because their memory is reclaimed, so using them afterwards may do
1081 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1082 one (it's much easier than Tom thought :-).
1084 (5) Because of (4), we can now allow for the use of $prvscheck_result
1085 inside the third argument.
1087 PH/03 For some reason, the default setting of PATH when running a command from
1088 a pipe transport was just "/usr/bin". I have changed it to
1091 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1092 anything to be listed in the output from -bV.
1094 PH/05 When a filter generated an autoreply, the entire To: header line was
1095 quoted in the delivery log line, like this:
1097 => >A.N.Other <ano@some.domain> <original@ddress> ...
1099 This has been changed so that it extracts the operative address. There
1100 may be more than one such address. If so, they are comma-separated, like
1103 => >ano@some.domain,ona@other.domain <original@ddress> ...
1105 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1106 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1107 not being looked up in the rDNS to get a host name, Exim was showing the
1108 IP address twice in Received: lines, even though the IP addresses were
1109 identical. For example:
1111 Received: from [1.2.3.4] (helo=[1.2.3.4])
1113 However, if the real host name was known, it was omitting the HELO data
1114 if it matched the actual IP address. This has been tidied up so that it
1115 doesn't show the same IP address twice.
1117 PH/07 When both +timestamp and +memory debugging was on, the value given by
1118 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1119 called by the debug printing, thereby overwriting the timestamp buffer.
1120 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1122 PH/08 When the original message was included in an autoreply transport, it
1123 always said "this is a copy of the message, including all the headers",
1124 even if body_only or headers_only was set. It now gives an appropriate
1127 PH/09 Applied a patch from the Sieve maintainer which:
1129 o fixes some comments
1130 o adds the (disabled) notify extension core
1131 o adds some debug output for the result of if/elsif tests
1132 o points to the current vacation draft in the documentation
1133 and documents the missing references header update
1137 o fixes a bug in processing the envelope test (when testing
1138 multiple envelope elements, the last element determinted the
1141 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1142 Electronic Mail") by including:
1144 Auto-submitted: auto-generated
1146 in the messages that it generates (bounce messages and others, such as
1147 warnings). In the case of bounce messages for non-SMTP mesages, there was
1148 also a typo: it was using "Auto_submitted" (underscore instead of
1149 hyphen). Since every message generated by Exim is necessarily in response
1150 to another message, thes have all been changed to:
1152 Auto-Submitted: auto-replied
1154 in accordance with these statements in the RFC:
1156 The auto-replied keyword:
1158 - SHOULD be used on messages sent in direct response to another
1159 message by an automatic process,
1161 - MUST NOT be used on manually-generated messages,
1163 - MAY be used on Delivery Status Notifications (DSNs) and Message
1164 Disposition Notifications (MDNs),
1166 - MUST NOT be used on messages generated by automatic or periodic
1167 processes, except for messages which are automatic responses to
1170 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1171 to the default Received: header definition.
1173 PH/12 Added log selector acl_warn_skipped (default on).
1175 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1176 variables because (a) they are in the wrong storage pool and (b) even if
1177 they were copied, it wouldn't work properly because of the caching.
1179 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1180 checking when decoding. Apparently there are clients that generate
1181 overlong encoded strings. Why am I not surprised?
1183 PH/15 If the first argument of "${if match_address" was not empty, but did not
1184 contain an "@" character, Exim crashed. Now it writes a panic log message
1185 and treats the condition as false.
1187 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1189 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1190 extension "envelope-auth". The code is finished and in agreement with
1191 other implementations, but there is no documentation so far and in fact,
1192 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1193 not changing the active code.
1195 Print executed "if" and "elsif" statements when debugging is used. This
1196 helps a great deal to understand what a filter does.
1198 Document more things not specified clearly in RFC3028. I had all this
1199 sorted out, when out of a sudden new issues came to my mind. Oops."
1201 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1204 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1205 canonical form (as documented). However, after a host name lookup from
1206 the IP address, check_host() was doing a simple string comparison with
1207 addresses acquired from the DNS when checking that the found name did
1208 have the original IP as one of its addresses. Since any found IPv6
1209 addresses are likely to be in abbreviated form, the comparison could
1210 fail. Luckily, there already exists a function for doing the comparison
1211 by converting both addresses to binary, so now that is used instead of
1212 the text comparison.
1214 PH/20 There was another similar case to PH/19, when a complete host name was
1215 given in a host list; looking up its IP address could give an abbreviated
1216 form, whereas the current host's name might or might not be abbreviated.
1217 The same fix has been applied.
1223 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1224 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1227 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1228 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1230 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1232 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1233 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1234 Also, for "Auto-Submitted", treat anything other than "no" as
1235 non-personal, in accordance with RFC 3834. (Previously it treated
1236 anything starting "auto-" as non-personal.)
1238 TF/01 The control=submission/name=... option had a problem with syntax
1239 errors if the name included a slash character. The /name= option
1240 now slurps the rest of the string, so it can include any characters
1241 but it must come last in the list of options (after /sender_retain
1244 PH/05 Some modifications to the interface to the fake nameserver for the new
1252 TK/01 Added the "success_on_redirect" address verification option. See
1253 NewStuff for rationale and an example.
1255 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1257 PH/02 Patch to exigrep to allow it to work on syslog lines.
1259 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1260 fread() to skip over the body file's header line, because in Cygwin the
1261 header line is locked and is inaccessible.
1263 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1264 co-exist for some time) to make it clear that it is the Exim ID that is
1265 referenced, not the Message-ID: header line.
1267 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1268 string_format() function, because snprintf() does not exist on all
1271 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1272 ${stat: expansion item.
1274 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1275 protocol synchronization error", to keep the pedants happy.
1277 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1278 well as for IRIX systems, when gcc is being used. See the host.c source
1281 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1283 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1286 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1287 transport and to the smtp transport in LMTP mode.
1289 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1291 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1292 run a filter in a subprocess. This could lead to confusion in subsequent
1293 lookups in the parent process. There should also be a search_tidyup() at
1294 the end of the subprocess.
1296 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1297 only if the host matched helo_try_verify_hosts, which caused the
1298 verification to occur when the EHLO/HELO command was issued. The ACL just
1299 tested the remembered result. Now, if a previous verification attempt has
1300 not happened, "verify = helo" does it there and then.
1302 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1304 TK/03 Fix log output including CR from clamd.
1306 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1307 exist provoked a memory error which could cause a segfault.
1309 PH/15 Installed PCRE 6.2
1311 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1313 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1314 of the problem. Specifically, suggested +O2 rather than +O1 for the
1317 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1319 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1320 record, the port was not correctly logged when the outgoing_port log
1321 selector was set (it logged the transort's default port).
1323 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1324 fallback_hosts, and "hosts" in the smtp transport.
1326 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1327 host errors such as "Connection refused".
1329 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1330 authentication with radiusclient 0.4.9:
1332 - Error returned from rc_read_config was caught wrongly
1333 - Username/password not passed on to radius server due to wrong length.
1335 The presumption is that some radiusclient API changes for 4.51/PH/17
1336 were not taken care of correctly. The code is still untested by me (my
1337 Linux distribution still has 0.3.2 of radiusclient), but it was
1338 contributed by a Radius user.
1340 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1341 expanding the "port" option of the smtp transport.
1343 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1344 while reading a MIME header. Thanks to Tom Hughes for a patch.
1346 PH/24 Include config.h inside local_scan.h so that configuration settings are
1349 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1350 that in an ACL for RCPT (for example), you can examine exactly what was
1353 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1354 commands, but it was not correctly comparing the address with the actual
1355 client host address. Thus, it would show the EHLO address in Received:
1356 header lines when this was not necessary.
1358 PH/27 Added the % operator to ${eval:}.
1360 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1361 it should be ignoring failures (because with -C, for example, it has lost
1362 privilege). It wasn't ignoring creation failures other than "already
1365 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1368 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1369 because an input line was too long, either on its own, or by virtue of
1370 too many continuations, the temporary file was not being removed, and the
1371 return code was incorrect.
1373 PH/31 Missing "BOOL" in function definition in filtertest.c.
1375 PH/32 Applied Sieve patches from the maintainer.
1377 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1379 PH/33 Added "verify = not_blind".
1381 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1382 Local/Makefile (with some defaults set). These are used in built scripts
1383 such as exicyclog, but they have never been used in the exim_install
1384 script (though there are many overriding facilities there). I have
1385 arranged that the exim_install script now takes note of these two
1388 PH/35 Installed configuration files for Dragonfly.
1390 PH/36 When a locally submitted message by a trusted user did not contain a
1391 From: header, and the sender address was obtained from -f or from an SMTP
1392 MAIL command, and the trusted user did not use -F to supply a sender
1393 name, $originator_name was incorrectly used when constructing a From:
1394 header. Furthermore, $originator_name was used for submission mode
1395 messages from external hosts without From: headers in a similar way,
1396 which is clearly wrong.
1398 PH/37 Added control=suppress_local_fixups.
1400 PH/38 When log_selector = +received_sender was set, and the addition of the
1401 sender made the log line's construction buffer exactly full, or one byte
1402 less than full, an overflow happened when the terminating "\n" was
1405 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1406 when the result of a list match is failure because a DNS lookup failed.
1408 PH/40 RM_COMMAND is now used in the building process.
1410 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1411 the "build-* directories that it finds.
1413 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1414 address in a domain literal was a prefix of an interface address.
1416 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1417 when verifying a sender address, unless rewrite_headers is false.
1419 PH/44 Wrote a long comment about why errors_to addresses are verified as
1420 recipients, not senders.
1422 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1423 the ratelimit ACL was added.
1425 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1427 PH/46 Added extra information about PostgreSQL errors to the error string.
1429 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1430 avoiding the need to install special zones in a real server. This is
1431 backwards compatible; if it can't find the fake resolver, it drops back.
1432 Thus, both old and new test suites can be run.
1434 TF/02 Added util/ratelimit.pl
1436 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1437 clock is set back in time.
1439 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1440 Candler <B.Candler@pobox.com>.
1442 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1443 OK for addresses that are the result of redirections.
1445 PH/48 A number of further additions for the benefit of the new test suite,
1446 including a fake gethostbyname() that interfaces to the fake DNS resolver
1449 TF/06 The fix for widen_domains has also been applied to qualify_single and
1450 search_parents which are the other dnslookup options that can cause
1453 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1456 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1458 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1459 service" if it encountered an MX record with an empty target hostname.
1460 The message is now "an MX or SRV record indicated no SMTP service".
1462 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1463 if the DNS of the sending site is misconfigured. This is quite a
1464 common situation. This change restores the behaviour of treating a
1465 helo verification defer as a failure.
1467 PH/51 If self=fail was set on a router, the bounce message did not include the
1468 actual error message.
1474 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1476 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1477 command itself ended in error, the underlying message about the transport
1478 filter timeout was being overwritten with the pipe command error. Now the
1479 underlying error message should be appended to the second error message.
1481 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1483 PH/02 Reduce the amount of output that "make" produces by default. Full output
1484 can still be requested.
1486 PH/03 The warning log line about a condition test deferring for a "warn" verb
1487 was being output only once per connection, rather than after each
1488 occurrence (because it was using the same function as for successful
1489 "warn" verbs). This seems wrong, so I have changed it.
1491 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1492 it should not have, which might have caused a crash in the right
1493 circumstances, but probably never did.
1495 PH/04 Installed a modified version of Tony Finch's patch to make submission
1496 mode fix the return path as well as the Sender: header line, and to
1497 add a /name= option so that you can make the user's friendly name appear
1500 TF/03 Added the control = fakedefer ACL modifier.
1502 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1503 Mark Lowes for thorough testing.
1505 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1507 TK/03 Merged latest SRS patch from Miles Wilton.
1509 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1510 with the definition in sysexits.h (which is #included earlier).
1511 Fortunately, Exim does not actually use EX_OK. The code used to try to
1512 preserve the sysexits.h value, by assumimg that macro definitions were
1513 scanned for macro replacements. I have been disabused of this notion,
1514 so now the code just undefines EX_OK before #including unistd.h.
1516 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1517 in the smtp transport. When a block could not be written in a single
1518 write() function, the timeout was being re-applied to each part-write.
1519 This seems wrong - if the receiver was accepting one byte at a time it
1520 would take for ever. The timeout is now adjusted when this happens. It
1521 doesn't have to be particularly precise.
1523 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1524 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1526 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1527 without a domain if the check to <postmaster@domain> fails.
1529 SC/01 Eximstats: added -xls and the ability to specify output files
1530 (patch written by Frank Heydlauf).
1532 SC/02 Eximstats: use FileHandles for outputing results.
1534 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1536 SC/04 Eximstats: fixed display of large numbers with -nvr option
1538 SC/05 Eximstats: fixed merging of reports with empty tables.
1540 SC/06 Eximstats: added the -include_original_destination flag
1542 SC/07 Eximstats: removed tabs and trailing whitespace.
1544 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1546 TK/06 MBOX spool code: Add real "From " MBOX separator line
1547 so the .eml file is really in mbox format (even though
1548 most programs do not really care). Patch from Alex Miller.
1550 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1551 The latter is generated from $received_to and is only set if the
1552 message has one envelope recipient. SA can use these headers,
1553 obviously out-of-the-box. Patch from Alex Miller.
1555 PH/08 The ${def test on a variable was returning false if the variable's
1556 value was "0", contrary to what the specification has always said!
1557 The result should be true unless the variable is empty.
1559 PH/09 The syntax error of a character other than { following "${if
1560 def:variable_name" (after optional whitespace) was not being diagnosed.
1561 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1562 accidental colon was present, for example, could give incorrect results.
1564 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1565 result is used (not including appendfile, where other changes are about
1568 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1569 This involved changing a lot of size variables from int to off_t. It
1570 should work with maildirs and everything.
1572 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1573 spamd dying while we are connected to it.
1575 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1578 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1579 function for BDB 4.3.
1581 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1583 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1584 the sources for PCRE have been re-arranged and more files are now
1587 PH/15 The code I had for printing potentially long long variables in PH/11
1588 above was not the best (it lost precision). The length of off_t variables
1589 is now inspected at build time, and an appropriate printing format (%ld
1590 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1591 to be "long long int" or "long int". This is needed for the internal
1592 formatting function string_vformat().
1594 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1595 the configuration file to be ":syslog", then the script "guesses" where
1596 the logs files are, rather than using the compiled in default. In our
1597 case the guess is not the same as the compiled default, so the script
1598 suddenly stopped working when I started to use syslog. The patch checks
1599 to see if log_file_path is "". If so, it attempts to read it from exim
1600 with no configuration file to get the compiled in version, before it
1601 falls back to the previous guessing code."
1603 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1604 implementing BATV in an Exim configuration. See NewStuff for the gory
1607 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1608 Makefile that are specific to HP-UX.
1610 PH/18 If the "use_postmaster" option was set for a recipient callout together
1611 with the "random" option, the postmaster address was used as the MAIL
1612 FROM address for the random test, but not for the subsequent recipient
1613 test. It is now used for both.
1615 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1616 patch removes a few documentation additions to RFC 3028, because the
1617 latest draft now contains them. It adds the new en;ascii-case comparator
1618 and a new error check for 8bit text in MIME parts. Comparator and
1619 require names are now matched exactly. I enabled the subaddress
1620 extension, but it is not well tested yet (read: it works for me)."
1622 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1623 rework some of the code of TK/09 above to avoid the hardwired use of
1624 "%lld" and "long long". Replaced the call to snprintf() with a call to
1627 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1628 records point to non-existent hosts", "retry timeout exceeded", and
1629 "retry time not reached for any host after a long failure period".
1631 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1632 experimental DomainKeys support:
1634 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1635 (2) On an error such as an illegally used "control", the wrong name for
1636 the control was given.
1638 These problems did NOT occur unless DomainKeys support was compiled.
1640 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1642 PH/24 Added ${if match_ip condition.
1644 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1645 databases so that it will be absolutely obvious if a crash occurs in the
1646 DB library. This is a regular occurrence (often caused by mis-matched
1649 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1650 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1651 picked up on a user's system that detects such things. There doesn't seem
1652 to be a gcc warning option for this - only an attribute that has to be
1653 put on the function's prototype. It seems that in Fedora Core 4 they have
1654 set this on a number of new functions. No doubt there will be more in due
1657 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1658 specify a transport. However, if an address that was verified by such a
1659 router was the subject of a callout, Exim crashed because it tried to
1660 read the rcpt_include_affixes from the non-existent transport. Now it
1661 just assumes that the setting of that option is false. This bug was
1662 introduced by 4.51/PH/31.
1664 PH/28 Changed -d+all to exclude +memory, because that information is very
1665 rarely of interest, but it makes the output a lot bigger. People tend to
1666 do -d+all out of habit.
1668 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1669 code in os-type was giving problems when libc.so lives in lib64, like on
1672 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1673 aren't the modern standard, and it seems that some systems' include files
1674 don't always have them. Exim was already checking for some of the newer
1675 ones like T_AAAA, and defining it itself. I've added checks for all the
1676 record types that Exim uses.
1678 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1679 not automatically generating a new one, as it is supposed to. This
1680 prevented TLS from working. If the file did exist, but contained invalid
1681 data, a new version was generated, as expected. It was only the case of a
1682 non-existent file that was broken.
1684 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1685 with a change in libdomainkeys > 0.64.
1687 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1688 from DNS. If the selector record carries the flag, it now has
1689 precedence over the domain-wide flag.
1691 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1693 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1694 the use of a port name that isn't defined in /etc/services) occurred, the
1695 message was deferred as in a normal delivery, and thus remained on the
1696 spool, instead of being failed because of the mua_wrapper setting. This
1697 is now fixed, and I tidied up some of the mua_wrapper messages at the
1700 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1701 the messages in a hash of arrays rather than using individual hashes.
1702 This is a bit cleaner and results in dramatic memory savings, albeit
1703 at a slight CPU cost.
1705 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1706 as requested by Marc Sherman.
1708 SC/10 Eximstats: added histograms for user specified patterns as requested
1711 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1713 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1714 fopen() in the content-scanning modules that did not already have it.
1720 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1721 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1723 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1725 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1727 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1728 to test Sieve filters that use "vacation".
1730 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1731 that changes the way the GnuTLS parameters are stored in the cache file.
1732 The new format can be generated externally. For backward compatibility,
1733 if the data in the cache doesn't make sense, Exim assumes it has read an
1734 old-format file, and it generates new data and writes a new file. This
1735 means that you can't go back to an older release without removing the
1738 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1739 work if there are any delivery delays because "one_time" forces the
1740 parent to be marked "delivered", so its unseen clone is never tried
1741 again. For this reason, Exim now forbids the simultaneous setting of
1744 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1745 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1746 entry if you want to know the details. The fix, however, neglected to
1747 consider the case when local delivery batching is involved. The test for
1748 "previously delivered" was not happening when checking to see if an
1749 address could be batched with a previous (undelivered) one; under
1750 certain circumstances this could lead to multiple deliveries to the same
1753 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1754 in its include files, and this causes problems building Exim.
1756 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1757 header_syntax) but Exim was just ignoring anything given after a slash.
1758 In particular, this caused confusion with an attempt to use "verify =
1759 reverse_host_lookup/defer_ok". An error is now given when options are
1760 supplied for verify items that do not have them. (Maybe reverse_host_
1761 lookup should have a defer_ok option, but that's a different point.)
1763 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1764 defined by RFC 821) to 2048, because there were problems with some AUTH
1765 commands, and RFC 1869 says the size should be increased for extended
1766 SMTP commands that take arguments.
1768 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1771 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1772 "unknown" error; now it says that the functionality isn't in the binary.
1774 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1775 an address' error message when a string expansion fails (syntax or
1776 whatever). Otherwise the password may appear in the log. Following change
1777 PH/42 below, there is no longer a chance of it appearing in a bounce
1780 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1782 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1783 of its addresses was ever tried. (Bugzilla bug #2.)
1785 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1786 the result incorrectly in the debug output. (It correctly added a newline
1787 to what was transported.)
1789 TF/01 Added $received_time.
1791 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1792 commented out examples of how to interface to a virus scanner and to
1793 SpamAssassin. Also added commented examples of av_scanner and
1794 spamd_address settings.
1796 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1797 and controls are allowed in which ACLs. There were a couple of minor
1798 errors. Some of the entries in the conditions table (which is a table of
1799 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1800 negation of where the condition IS allowed.
1802 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1804 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1805 header file does not have a version number, so I've had to invent a new
1806 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1807 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1808 radiusclient), but it was contributed by a Radius user.
1810 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1811 files or directories, for OpenSSL.
1813 PH/19 When an Exim process that is running as root has to create an Exim log
1814 file, it does so in a subprocess that runs as exim:exim so as to get the
1815 ownership right at creation (otherwise, other Exim processes might see
1816 the file with the wrong ownership). There was no test for failure of this
1817 fork() call, which would lead to the process getting stuck as it waited
1818 for a non-existent subprocess. Forks do occasionally fail when resources
1819 run out. I reviewed all the other calls to fork(); they all seem to check
1822 PH/20 When checking for unexpected SMTP input at connect time (before writing
1823 the banner), Exim was not dealing correctly with a non-positive return
1824 from the read() function. If the client had disconnected by this time,
1825 the result was a log entry for a synchronization error with an empty
1826 string after "input=" when read() returned zero. If read() returned -1
1827 (an event I could not check), uninitialized data bytes were printed.
1828 There were reports of junk text (parts of files, etc) appearing after
1831 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1833 PH/22 Added support for macro redefinition, and (re)definition in between
1834 driver and ACL definitions.
1836 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1837 forgetting to use the resulting value; it was using the unexpanded value.
1839 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1840 hadn't been configured. The fix is from Juergen Kreileder, who
1841 understands it better than I do:
1843 "Here's what I see happening with three configured cyrus_sasl
1844 authenticators configured (plain, login, cram-md5):
1846 On startup auth_cyrus_sasl_init() gets called for each of these.
1847 This means three calls to sasl_listmech() without a specified mech_list.
1848 => SASL tests which mechs of all available mechs actually work
1849 => three warnings about OTP not working
1850 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1852 With the patch, sasl_listmech() also gets called three times. But now
1853 SASL's mech_list option is set to the server_mech specified in the the
1854 authenticator. Or in other words, the answer from sasl_listmech()
1855 gets limited to just the mech you're testing for (which is different
1857 => the return list contains just 'plain' or 'login', 'cram-md5' or
1858 nothing depending on the value of ob->server_mech.
1860 I've just tested the patch: Authentication still works fine,
1861 unavailable mechs specified in the exim configuration are still
1862 caught, and the auth.log warnings about OTP are gone."
1864 PH/25 When debugging is enabled, the contents of the command line are added
1865 to the debugging output, even when log_selector=+arguments is not
1868 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1869 answer is "GNU", and only if the return is "GNU/something" is the answer
1872 PH/27 $acl_verify_message is now set immediately after the failure of a
1873 verification in an ACL, and so is available in subsequent modifiers. In
1874 particular, the message can be preserved by coding like this:
1876 warn !verify = sender
1877 set acl_m0 = $acl_verify_message
1879 Previously, $acl_verify_message was set only while expanding "message"
1880 and "log_message" when a very denied access.
1882 PH/28 Modified OS/os.c-Linux with
1884 -#ifndef OS_LOAD_AVERAGE
1885 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1887 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1888 nomenclature these days.)
1890 PH/29 Installed patch from the Sieve maintainer that adds the options
1891 sieve_useraddress and sieve_subaddress to the redirect router.
1893 PH/30 In these circumstances:
1894 . Two addresses routed to the same list of hosts;
1895 . First host does not offer TLS;
1896 . First host accepts first address;
1897 . First host gives temporary error to second address;
1898 . Second host offers TLS and a TLS session is established;
1899 . Second host accepts second address.
1900 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1901 and peerdn, if requested) that were in fact used only for the second
1904 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1905 attention to any local part prefix or suffix that was matched by the
1906 router that accepted the address. It now behaves in the same way as it
1907 does for delivery: the affixes are removed from the local part unless
1908 rcpt_include_affixes is set on the transport.
1910 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1911 timeout during the DATA phase of an incoming message.
1913 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1914 applied a patch sanctioned by the Sieve maintainer.
1916 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1917 the uid or gid is negative. A case of a negative gid caused this to be
1918 noticed. The fix allows for either to be negative.
1920 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1921 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1924 PH/36 The change PH/12 above was broken. Fixed it.
1926 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1927 the grounds that routing the same address twice would always produce the
1928 same answer. This might have been true once, but it is certainly no
1929 longer true now. Routing a child address may depend on the previous
1930 routing that produced that child. Some complicated redirection strategies
1931 went wrong when messages had multiple recipients, and made Exim's
1932 behaviour dependent on the order in which the addresses were given.
1934 I have moved the duplicate checking until after the routing is complete.
1935 Exim scans the addresses that are assigned to local and remote
1936 transports, and removes any duplicates. This means that more work will be
1937 done, as duplicates will always all be routed, but duplicates are
1938 presumably rare, so I don't expect this is of any significance.
1940 For deliveries to pipes, files, and autoreplies, the duplicate checking
1941 still happens during the routing process, since they are not going to be
1944 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1945 It corrects a timeout issue with spamd. This is Ian's comment: "The
1946 background is that sometimes spamd either never reads data from a
1947 connection it has accepted, or it never writes response data. The exiscan
1948 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1949 blindly assumes that writes won't block so it may never time out."
1951 PH/39 Allow G after quota size as well as K and M.
1953 PH/40 The value set for $authenticated_id in an authenticator may not contain
1954 binary zeroes or newlines because the value is written to log lines and
1955 to spool files. There was no check on this. Now the value is run through
1956 the string_printing() function so that such characters are converted to
1957 printable escape sequences.
1959 PH/41 $message_linecount is a new variable that contains the total number of
1960 lines in the message. Compare $body_linecount, which is the count for the
1963 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1964 bounce and delay warning messages, except in certain special cases, which
1967 (a) An SMTP error message from a remote host;
1968 (b) A message specified in a :fail: redirection;
1969 (c) A message specified in a "fail" command in a system filter;
1970 (d) A message specified in a FAIL return from the queryprogram router;
1971 (e) A message specified by the cannot_route_message router option.
1973 In these cases only, Exim does include the error details in bounce and
1974 warning messages. There are also a few cases where bland messages such
1975 as "unrouteable address" or "local delivery error" are given.
1977 PH/43 $value is now also set for the "else" part of a ${run expansion.
1979 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1980 being worked on, but at least Exim now implements the latest version to
1983 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1984 process to complete was treated as a delivery failure, a timeout while
1985 writing the message to the pipe was logged, but erroneously treated as a
1986 successful delivery. Such timeouts include transport filter timeouts. For
1987 consistency with the overall process timeout, these timeouts are now
1988 treated as errors, giving rise to delivery failures by default. However,
1989 there is now a new Boolean option for the pipe transport called
1990 timeout_defer, which, if set TRUE, converts the failures into defers for
1991 both kinds of timeout. A transport filter timeout is now identified in
1994 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1995 systems where "make" and "gmake" are different, calling "gmake" at top
1996 level broke things. I've arranged for the value of $(MAKE) to be passed
1997 from the Makefile to this script so that it can call the same version of
2001 A note about Exim versions 4.44 and 4.50
2002 ----------------------------------------
2004 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2005 changes of various kinds. As a consequence, a big documentation update was
2006 needed. This delayed the release for rather longer than seemed good, especially
2007 in the light of a couple of (minor) security issues. Therefore, the changes
2008 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2009 release. So 4.44 and 4.50 are in effect two different branches that both start
2012 I have left the 4.50 change log unchanged; it contains all the changes since
2013 4.43. The change log for 4.44 is below; many of its items are identical to
2014 those for 4.50. This seems to be the most sensible way to preserve the
2015 historical information.
2021 1. Minor wording change to the doc/README.SIEVE file.
2023 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2024 computation of the current number of files was incorrect.
2026 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2027 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2030 4. Give more explanation in the error message when the command for a transport
2031 filter fails to execute.
2033 5. There are several places where Exim runs a non-Exim command in a
2034 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2035 was being done only for the command run by the queryprogram router. It is
2036 now done for all such subprocesses. The other cases are: ${run, transport
2037 filters, and the commands run by the lmtp and pipe transports.
2039 6. Added CONFIGURE_GROUP build-time option.
2041 7. Some older OS have a limit of 256 on the maximum number of file
2042 descriptors. Exim was using setrlimit() to set 1000 as a large value
2043 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2044 systems. I've change it so that if it can't get 1000, it tries for 256.
2046 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2047 was an oversight, and furthermore, ever since the addition of extra
2048 controls (e.g. 4.43/32), the checks on when to allow different forms of
2049 "control" were broken. There should now be diagnostics for all cases when a
2050 control that does not make sense is encountered.
2052 9. Added the /retain_sender option to "control=submission".
2054 10. $recipients is now available in the predata ACL (oversight).
2056 11. Tidy the search cache before the fork to do a delivery from a message
2057 received from the command line. Otherwise the child will trigger a lookup
2058 failure and thereby defer the delivery if it tries to use (for example) a
2059 cached ldap connection that the parent has called unbind on.
2061 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2062 of $address_data from the recipient verification was clobbered by the
2063 sender verification.
2065 13. The value of address_data from a sender verification is now available in
2066 $sender_address_data in subsequent conditions in the ACL statement.
2068 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2070 15. Added a new option "connect=<time>" to callout options, to set a different
2073 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2074 was its contents. (It was OK if the option was not defined at all.)
2076 17. A "Completed" log line is now written for messages that are removed from
2077 the spool by the -Mrm option.
2079 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2080 information about exactly what failed.
2082 19. Added -dd to debug only the daemon process.
2084 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2085 handles timeouts, both on the server side and network timeouts. Renamed the
2086 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2088 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2089 It is now set to "smtps".
2091 22. $host_address is now set to the target address during the checking of
2092 ignore_target_hosts.
2094 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2095 being passed; this would have caused $sender_host_name to have been used if
2096 matching the list had actually called for a host name (not very likely,
2097 since this list is usually IP addresses). A host name is now passed as
2100 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2101 code that specifies a non-restarting handler (typically sigaction() in
2102 modern systems) in an attempt to fix a rare and obscure crash bug.
2104 25. Narrowed the window for a race in the daemon that could cause it to ignore
2105 SIGCHLD signals. This is not a major problem, because they are used only to
2106 wake it up if nothing else does.
2108 26. A malformed maildirsize file could cause Exim to calculate negative values
2109 for the mailbox size or file count. Odd effects could occur as a result.
2110 The maildirsize information is now recalculated if the size or filecount
2113 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2114 support for a long time. Removed HAVE_SYS_VFS_H.
2116 28. Installed the latest version of exipick from John Jetmore.
2118 29. In an address list, if the pattern was not a regular expression, an empty
2119 subject address (from a bounce message) matched only if the pattern was an
2120 empty string. Non-empty patterns were not even tested. This was the wrong
2121 because it is perfectly reasonable to use an empty address as part of a
2122 database query. An empty address is now tested by patterns that are
2123 lookups. However, all the other forms of pattern expect the subject to
2124 contain a local part and a domain, and therefore, for them, an empty
2125 address still always fails if the pattern is not itself empty.
2127 30. Exim went into a mad DNS loop when attempting to do a callout where the
2128 host was specified on an smtp transport, and looking it up yielded more
2129 than one IP address.
2131 31. Re-factored the code for checking spool and log partition space into a
2132 function that finds that data and another that does the check. The former
2133 is then used to implement four new variables: $spool_space, $log_space,
2134 $spool_inodes, and $log_inodes.
2136 32. The RFC2047 encoding function was originally intended for short strings
2137 such as real names; it was not keeping to the 75-character limit for
2138 encoded words that the RFC imposes. It now respects the limit, and
2139 generates multiple encoded words if necessary. To be on the safe side, I
2140 have increased the buffer size for the ${rfc2047: expansion operator from
2143 33. It is now permitted to omit both strings after an "if" condition; if the
2144 condition is true, the result is "true". As before, when the second string
2145 is omitted, a false condition yields an empty string. This makes it less
2146 cumbersome to write custom ACL and router conditions.
2148 34. Failure to deliver a bounce message always caused it to be frozen, even if
2149 there was an errors_to setting on the router. The errors_to setting is now
2152 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2153 canonical form (fully expanded) before being placed in
2154 $sender_host_address.
2156 36. The table in the code that translates DNS record types into text (T_A to
2157 "A" for instance) was missing entries for NS and CNAME. It is just possible
2158 that this could have caused confusion if both these types were looked up
2159 for the same domain, because the text type is used as part of Exim's
2160 per-process caching. But the chance of anyone hitting this buglet seems
2163 37. The dnsdb lookup has been extended in a number of ways.
2165 (1) There is a new type, "zns", which walks up the domain tree until it
2166 finds some nameserver records. It should be used with care.
2168 (2) There is a new type, "mxh", which is like "mx" except that it returns
2169 just the host names, not the priorities.
2171 (3) It is now possible to give a list of domains (or IP addresses) to be
2172 looked up. The behaviour when one of the lookups defers can be
2173 controlled by a keyword.
2175 (4) It is now possible to specify the separator character for use when
2176 multiple records are returned.
2178 38. The dnslists ACL condition has been extended: it is now possible to supply
2179 a list of IP addresses and/or domains to be looked up in a particular DNS
2182 39. Added log_selector=+queue_time_overall.
2184 40. When running the queue in the test harness, wait just a tad after forking a
2185 delivery process, to get repeatability of debugging output.
2187 41. Include certificate and key file names in error message when GnuTLS fails
2188 to set them up, because the GnuTLS error message doesn't include the name
2189 of the failing file when there is a problem reading it.
2191 42. Allow both -bf and -bF in the same test run.
2193 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2195 44. The "Exiscan patch" is now merged into the mainline Exim source.
2197 45. Sometimes the final signoff response after QUIT could fail to get
2198 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2199 < 0 before doing a fflush(). This bug looks as though it goes back to the
2200 introduction of TLS in release 3.20, but "sometimes" must have been rare
2201 because the tests only now provoked it.
2203 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2204 (this can affect the format of dates).
2206 47. exim_tidydb, when checking for the continued existence of a message for
2207 which it has found a message-specific retry record, was not finding
2208 messages that were in split spool directories. Consequently, it was
2209 deleting retry records that should have stayed in existence.
2211 48. Steve fixed some bugs in eximstats.
2213 49. The SPA authentication driver was not abandoning authentication and moving
2214 on to the next authenticator when an expansion was forced to fail,
2215 contradicting the general specification for all authenticators. Instead it
2216 was generating a temporary error. It now behaves as specified.
2218 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2219 (the order specifies the preference for clients). The order is now AES256,
2220 AES128, 3DES, ARCFOUR128.
2222 51. Small patch to Sieve code - explicitly set From: when generating an
2225 52. Exim crashed if a remote delivery caused a very long error message to be
2226 recorded - for instance if somebody sent an entire SpamAssassin report back
2227 as a large number of 550 error lines. This bug was coincidentally fixed by
2228 increasing the size of one of Exim's internal buffers (big_buffer) that
2229 happened as part of the Exiscan merge. However, to be on the safe side, I
2230 have made the code more robust (and fixed the comments that describe what
2233 53. Now that there can be additional text after "Completed" in log lines (if
2234 the queue_time_overall log selector is set), a one-byte patch to exigrep
2235 was needed to allow it to recognize "Completed" as not the last thing in
2238 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2239 patch that reportedly fixes this has been added. I am not expert enough to
2240 create a test for it. This is what the patch creator wrote:
2242 "I found a little strange behaviour of ldap code when working with
2243 Windows 2003 AD Domain, where users was placed in more than one
2244 Organization Units. When I tried to give exim partial DN, the exit code
2245 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2246 But simultaneously result of request was absolutely normal ldap result,
2247 so I produce this patch..."
2249 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2250 REFERENCE, so I have modified the code to exclude the patch when that macro
2253 55. Some experimental protocols are using DNS PTR records for new purposes. The
2254 keys for these records are domain names, not reversed IP addresses. The
2255 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2256 leaves it alone. Component reversal etc. now happens only for IP addresses.
2259 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2261 57. Double the size of the debug message buffer (to 2048) so that more of very
2262 long debug lines gets shown.
2264 58. The exicyclog utility now does better if the number of log files to keep
2265 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2267 59. Two changes related to the smtp_active_hostname option:
2269 (1) $smtp_active_hostname is now available as a variable.
2270 (2) The default for smtp_banner uses $smtp_active_hostname instead
2271 of $primary_hostname.
2273 60. The host_aton() function is supposed to be passed a string that is known
2274 to be a valid IP address. However, in the case of IPv6 addresses, it was
2275 not checking this. This is a hostage to fortune. Exim now panics and dies
2276 if the condition is not met. A case was found where this could be provoked
2277 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2278 components; fortuitously, this particular loophole had already been fixed
2279 by change 4.50/55 above.
2281 If there are any other similar loopholes, the new check in host_aton()
2282 itself should stop them being exploited. The report I received stated that
2283 data on the command line could provoke the exploit when Exim was running as
2284 exim, but did not say which command line option was involved. All I could
2285 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2286 running as the user.
2289 61. There was a buffer overflow vulnerability in the SPA authentication code
2290 (which came originally from the Samba project). I have added a test to the
2291 spa_base64_to_bits() function which I hope fixes it.
2294 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2295 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2297 63. The daemon start-up calls getloadavg() while still root for those OS that
2298 need the first call to be done as root, but it missed one case: when
2299 deliver_queue_load_max is set with deliver_drop_privilege. This is
2300 necessary for the benefit of the queue runner, because there is no re-exec
2301 when deliver_drop_privilege is set.
2303 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2304 This has been fixed.
2306 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2307 was in use, was not putting the data itself into the right store pool;
2308 consequently, it could be overwritten for a subsequent message in the same
2309 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2312 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2314 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2315 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2316 the function were treating the return as a boolean value, which happened to
2317 work because 0=false and not-0=true, but is not correct code.
2319 68. The host_aton() function was not handling scoped IPv6 addresses (those
2320 with, for example, "%eth0" on the end) correctly.
2322 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2323 negated items (that is, ~something) in unsigned ints. Some compilers
2324 apparently mutter when there is no cast.
2326 70. If an address verification called from an ACL failed, and did not produce a
2327 user-specific message (i.e. there was only a "system" message), nothing was
2328 put in $acl_verify_message. In this situation, it now puts the system
2331 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2332 session; change 4.31/43 added the unwanted input to the log line - except
2333 that it did not do this in the start of session case. It now does.
2335 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2336 This is wrong and can cause the other end to generate a synchronization
2337 error if it is another Exim or anything else that does the synchronization
2338 check. A QUIT command is no longer sent after a timeout.
2340 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2341 during host lookups.
2343 74. The defer_ok option of callout verification was not working if it was used
2344 when verifying addresses in header lines, that is, for this case:
2346 verify = header_sender/callout=defer_ok
2348 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2349 those file descriptors could be used for SMTP connections. If anything
2350 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2351 could be sent to the SMTP client, causing chaos. The daemon now opens
2352 stdin, stdout, and stderr to /dev/null when it puts itself into the
2355 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2356 log by default. The user can override this with suitable Perl magic.
2358 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2359 the log message when discard triggers, was not working for the DATA ACL or
2360 for the non-SMTP ACL.
2362 78. Error message wording change in sieve.c.
2364 79. If smtp_accept_max_per_host was set, the number of connections could be
2365 restricted to fewer than expected, because the daemon was trying to set up
2366 a new connection before checking whether the processes handling previous
2367 connections had finished. The check for completed processes is now done
2368 earlier. On busy systems, this bug wouldn't be noticed because something
2369 else would have woken the daemon, and it would have reaped the completed
2372 80. If a message was submitted locally by a user whose login name contained one
2373 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2374 It caused a spool format error. I have fixed the spool reading code. A
2375 related problem was that the "from" clause in the Received: line became
2376 illegal because of the space(s). It is now covered by ${quote_local_part.
2378 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2381 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2383 83. Patch from Sieve maintainer to fix unterminated string problem in
2384 "vacation" handling.
2386 84. Some minor changes to the Linux configuration files to help with other
2387 OS variants using glibc.
2389 85. One more patch for Sieve to update vacation handling to latest spec.
2392 ----------------------------------------------------
2393 See the note above about the 4.44 and 4.50 releases.
2394 ----------------------------------------------------
2400 1. Change 4.43/35 introduced a bug that caused file counts to be
2401 incorrectly computed when quota_filecount was set in an appendfile
2404 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2405 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2408 3. Give more explanation in the error message when the command for a transport
2409 filter fails to execute.
2411 4. There are several places where Exim runs a non-Exim command in a
2412 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2413 was being done only for the command run by the queryprogram router. It is
2414 now done for all such subprocesses. The other cases are: ${run, transport
2415 filters, and the commands run by the lmtp and pipe transports.
2417 5. Some older OS have a limit of 256 on the maximum number of file
2418 descriptors. Exim was using setrlimit() to set 1000 as a large value
2419 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2420 systems. I've change it so that if it can't get 1000, it tries for 256.
2422 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2423 was an oversight, and furthermore, ever since the addition of extra
2424 controls (e.g. 4.43/32), the checks on when to allow different forms of
2425 "control" were broken. There should now be diagnostics for all cases when a
2426 control that does not make sense is encountered.
2428 7. $recipients is now available in the predata ACL (oversight).
2430 8. Tidy the search cache before the fork to do a delivery from a message
2431 received from the command line. Otherwise the child will trigger a lookup
2432 failure and thereby defer the delivery if it tries to use (for example) a
2433 cached ldap connection that the parent has called unbind on.
2435 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2436 of $address_data from the recipient verification was clobbered by the
2437 sender verification.
2439 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2440 was its contents. (It was OK if the option was not defined at all.)
2442 11. A "Completed" log line is now written for messages that are removed from
2443 the spool by the -Mrm option.
2445 12. $host_address is now set to the target address during the checking of
2446 ignore_target_hosts.
2448 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2449 being passed; this would have caused $sender_host_name to have been used if
2450 matching the list had actually called for a host name (not very likely,
2451 since this list is usually IP addresses). A host name is now passed as
2454 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2455 code that specifies a non-restarting handler (typically sigaction() in
2456 modern systems) in an attempt to fix a rare and obscure crash bug.
2458 15. Narrowed the window for a race in the daemon that could cause it to ignore
2459 SIGCHLD signals. This is not a major problem, because they are used only to
2460 wake it up if nothing else does.
2462 16. A malformed maildirsize file could cause Exim to calculate negative values
2463 for the mailbox size or file count. Odd effects could occur as a result.
2464 The maildirsize information is now recalculated if the size or filecount
2467 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2468 support for a long time. Removed HAVE_SYS_VFS_H.
2470 18. Updated exipick to current release from John Jetmore.
2472 19. Allow an empty sender to be matched against a lookup in an address list.
2473 Previously the only cases considered were a regular expression, or an
2476 20. Exim went into a mad DNS lookup loop when doing a callout where the
2477 host was specified on the transport, if the DNS lookup yielded more than
2480 21. The RFC2047 encoding function was originally intended for short strings
2481 such as real names; it was not keeping to the 75-character limit for
2482 encoded words that the RFC imposes. It now respects the limit, and
2483 generates multiple encoded words if necessary. To be on the safe side, I
2484 have increased the buffer size for the ${rfc2047: expansion operator from
2487 22. Failure to deliver a bounce message always caused it to be frozen, even if
2488 there was an errors_to setting on the router. The errors_to setting is now
2491 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2492 canonical form (fully expanded) before being placed in
2493 $sender_host_address.
2495 24. Updated eximstats to version 1.33
2497 25. Include certificate and key file names in error message when GnuTLS fails
2498 to set them up, because the GnuTLS error message doesn't include the name
2499 of the failing file when there is a problem reading it.
2501 26. Expand error message when OpenSSL has problems setting up cert/key files.
2504 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2505 (this can affect the format of dates).
2507 28. exim_tidydb, when checking for the continued existence of a message for
2508 which it has found a message-specific retry record, was not finding
2509 messages that were in split spool directories. Consequently, it was
2510 deleting retry records that should have stayed in existence.
2512 29. eximstats updated to version 1.35
2513 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2514 1.35 - bugfix such that pie charts by volume are generated correctly
2516 30. The SPA authentication driver was not abandoning authentication and moving
2517 on to the next authenticator when an expansion was forced to fail,
2518 contradicting the general specification for all authenticators. Instead it
2519 was generating a temporary error. It now behaves as specified.
2521 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2522 (the order specifies the preference for clients). The order is now AES256,
2523 AES128, 3DES, ARCFOUR128.
2525 31. Small patch to Sieve code - explicitly set From: when generating an
2528 32. Exim crashed if a remote delivery caused a very long error message to be
2529 recorded - for instance if somebody sent an entire SpamAssassin report back
2530 as a large number of 550 error lines. This bug was coincidentally fixed by
2531 increasing the size of one of Exim's internal buffers (big_buffer) that
2532 happened as part of the Exiscan merge. However, to be on the safe side, I
2533 have made the code more robust (and fixed the comments that describe what
2536 33. Some experimental protocols are using DNS PTR records for new purposes. The
2537 keys for these records are domain names, not reversed IP addresses. The
2538 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2539 leaves it alone. Component reversal etc. now happens only for IP addresses.
2542 34. The host_aton() function is supposed to be passed a string that is known
2543 to be a valid IP address. However, in the case of IPv6 addresses, it was
2544 not checking this. This is a hostage to fortune. Exim now panics and dies
2545 if the condition is not met. A case was found where this could be provoked
2546 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2547 components; fortuitously, this particular loophole had already been fixed
2548 by change 4.50/55 or 4.44/33 above.
2550 If there are any other similar loopholes, the new check in host_aton()
2551 itself should stop them being exploited. The report I received stated that
2552 data on the command line could provoke the exploit when Exim was running as
2553 exim, but did not say which command line option was involved. All I could
2554 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2555 running as the user.
2558 35. There was a buffer overflow vulnerability in the SPA authentication code
2559 (which came originally from the Samba project). I have added a test to the
2560 spa_base64_to_bits() function which I hope fixes it.
2563 36. The daemon start-up calls getloadavg() while still root for those OS that
2564 need the first call to be done as root, but it missed one case: when
2565 deliver_queue_load_max is set with deliver_drop_privilege. This is
2566 necessary for the benefit of the queue runner, because there is no re-exec
2567 when deliver_drop_privilege is set.
2569 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2570 was in use, was not putting the data itself into the right store pool;
2571 consequently, it could be overwritten for a subsequent message in the same
2572 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2575 38. Sometimes the final signoff response after QUIT could fail to get
2576 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2577 < 0 before doing a fflush(). This bug looks as though it goes back to the
2578 introduction of TLS in release 3.20, but "sometimes" must have been rare
2579 because the tests only now provoked it.
2585 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2586 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2587 Now it returns an integer. A number of calls were still expecting a T/F
2588 return. Fortuitously, in all cases, the tests worked in OK situations,
2589 which is the norm. However, things would have gone wrong on any write
2590 failures on the smtp file descriptor. This function is used when sending
2591 messages over SMTP and also when doing verify callouts.
2593 2. When Exim is called to do synchronous delivery of a locally submitted
2594 message (the -odf or -odi options), it no longer closes stderr before doing
2597 3. Implemented the mua_wrapper option.
2599 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2601 5. Implemented the functions header_remove(), header_testname(),
2602 header_add_at_position(), and receive_remove_recipient(), and exported them
2605 6. If an ACL "warn" statement specified the addition of headers, Exim already
2606 inserted X-ACL-Warn: at the start if there was no header name. However, it
2607 was not making this test for the second and subsequent header lines if
2608 there were newlines in the string. This meant that an invalid header could
2609 be inserted if Exim was badly configured.
2611 7. Allow an ACL "warn" statement to add header lines at the start or after all
2612 the Received: headers, as well as at the end.
2614 8. Added the rcpt_4xx retry error code.
2616 9. Added postmaster_mailfrom=xxx to callout verification option.
2618 10. Added mailfrom=xxxx to the callout verification option, for verify=
2621 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2622 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2624 12. Inserted some casts to stop certain compilers warning when using pointer
2625 differences as field lengths or precisions in printf-type calls (mostly
2626 affecting debugging statements).
2628 13. Added optional readline() support for -be (dynamically loaded).
2630 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2631 same clock tick as a message's arrival, so that its received time was the
2632 same as the "first fail" time on the retry record, and that message
2633 remained on the queue past the ultimate address timeout, every queue runner
2634 would try a delivery (because it was past the ultimate address timeout) but
2635 after another failure, the ultimate address timeout, which should have then
2636 bounced the address, did not kick in. This was a "< instead of <=" error;
2637 in most cases the first failure would have been in the next clock tick
2638 after the received time, and all would be well.
2640 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2641 being recognized when the domain list was tested by the match_domain
2642 condition in an expansion string.
2644 16. Added the ${str2b64: operator.
2646 17. Exim was always calling setrlimit() to set a large limit for the number of
2647 processes, without checking whether the existing limit was already
2648 adequate. (It did check for the limit on file descriptors.) Furthermore,
2649 errors from getrlimit() and setrlimit() were being ignored. Now they are
2650 logged to the main and panic logs, but Exim does carry on, to try to do its
2651 job under whatever limits there are.
2653 18. Imported PCRE 5.0.
2655 19. Trivial typo in log message " temporarily refused connection" (the leading
2658 20. If the log selector return_path_on_delivery was set and an address was
2659 redirected to /dev/null, the delivery process crashed because it assumed
2660 that a return path would always be set for a "successful" delivery. In this
2661 case, the whole delivery is bypassed as an optimization, and therefore no
2664 21. Internal re-arrangement: the function for sending a challenge and reading
2665 a response while authentication was assuming a zero-terminated challenge
2666 string. It's now changed to take a pointer and a length, to allow for
2667 binary data in such strings.
2669 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2671 23. Exim was not respecting finduser_retries when seeking the login of the
2672 uid under which it was called; it was always trying 10 times. (The default
2673 setting of finduser_retries is zero.) Also, it was sleeping after the final
2674 failure, which is pointless.
2676 24. Implemented tls_on_connect_ports.
2678 25. Implemented acl_smtp_predata.
2680 26. If the domain in control=submission is set empty, Exim assumes that the
2681 authenticated id is a complete email address when it generates From: or
2682 Sender: header lines.
2684 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2685 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2686 chown and chgrp in /bin and hostname in /usr/bin.
2688 28. Exim was keeping the "process log" file open after each use, just as it
2689 does for the main log. This opens the possibility of it remaining open for
2690 long periods when the USR1 signal hits a daemon. Occasional processlog
2691 errors were reported, that could have been caused by this. Anyway, it seems
2692 much more sensible not to leave this file open at all, so that is what now
2695 29. The long-running daemon process does not normally write to the log once it
2696 has entered its main loop, and it closes the log before doing so. This is
2697 so that log files can straightforwardly be renamed and moved. However,
2698 there are a couple of unusual error situations where the daemon does write
2699 log entries, and I had neglected to close the log afterwards.
2701 30. The text of an SMTP error response that was received during a remote
2702 delivery was being truncated at 512 bytes. This is too short for some of
2703 the long messages that one sometimes sees. I've increased the limit to
2706 31. It is now possible to make retry rules that apply only when a message has a
2707 specific sender, in particular, an empty sender.
2709 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2710 it possible to be selective about when SMTP synchronization is enforced.
2712 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2714 32. Implemented hosts_connection_nolog.
2716 33. Added an ACL for QUIT.
2718 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2721 35. Added mailbox_size and mailbox_filecount to appendfile.
2723 36. Added control = no_multiline_responses to ACLs.
2725 37. There was a bug in the logic of the code that waits for the clock to tick
2726 in the case where the clock went backwards by a substantial amount such
2727 that the microsecond fraction of "now" was more than the microsecond
2728 fraction of "then" (but the whole seconds number was less).
2730 38. Added support for the libradius Radius client library this is found on
2731 FreeBSD (previously only the radiusclient library was supported).
2737 1. When certain lookups returned multiple values in the form name=value, the
2738 quoting of the values was not always being done properly. Specifically:
2739 (a) If the value started with a double quote, but contained no whitespace,
2741 (b) If the value contained whitespace other than a space character (i.e.
2742 tabs or newlines or carriage returns) it was not quoted.
2743 This fix has been applied to the mysql and pgsql lookups by writing a
2744 separate quoting function and calling it from the lookup code. The fix
2745 should probably also be applied to nisplus, ibase and oracle lookups, but
2746 since I cannot test any of those, I have not disturbed their existing code.
2748 2. A hit in the callout cache for a specific address caused a log line with no
2749 reason for rejecting RCPT. Now it says "Previous (cached) callout
2750 verification failure".
2752 3. There was an off-by-one bug in the queryprogram router. An over-long
2753 return line was truncated at 256 instead of 255 characters, thereby
2754 overflowing its buffer with the terminating zero. As well as fixing this, I
2755 have increased the buffer size to 1024 (and made a note to document this).
2757 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2758 when Exim is waiting for an SMTP response from a remote server, Exim
2759 restarts its select() call on the socket, thereby resetting its timeout.
2760 This is not a problem when such interrupts are rare. Somebody set up a cron
2761 job to run exiwhat every 2 minutes, which is less than the normal select()
2762 timeout (5 or 10 minutes). This meant that the select() timeout never
2763 kicked in because it was always reset. I have fixed this by comparing the
2764 time when an interrupt arrives with the time at the start of the first call
2765 to select(). If more time than the timeout has elapsed, the interrupt is
2766 treated as a timeout.
2768 5. Some internal re-factoring in preparation for the addition of Sieve
2769 extensions (by MH). In particular, the "personal" test is moved to a
2770 separate function, and given an option for scanning Cc: and Bcc: (which is
2771 not set for Exim filters).
2773 6. When Exim created an email address using the login of the caller as the
2774 local part (e.g. when creating a From: or Sender: header line), it was not
2775 quoting the local part when it contained special characters such as @.
2777 7. Installed new OpenBSD configuration files.
2779 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2780 try to make them clearer.
2782 9. Callout options, other than the timeout value, were being ignored when
2783 verifying sender addresses in header lines. For example, when using
2785 verify = header_sender/callout=no_cache
2787 the cache was (incorrectly) being used.
2789 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2790 only the Cygwin environment.
2792 11. When return_path_on_delivery was set as a log selector, if different remote
2793 addresses in the same message used different return paths and parallel
2794 remote delivery occurred, the wrong values would sometimes be logged.
2795 (Whenever a remote delivery process finished, the return path value from
2796 the most recently started remote delivery process was logged.)
2798 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2799 lines when AUTH and/or TLS are in use. This is the "received protocol"
2800 field. Exim used to use "asmtp" for authenticated SMTP, without any
2801 indication (in the protocol name) for TLS use. Now it follows the RFC and
2802 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2803 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2804 names appear in log lines as well as in Received: header lines.
2806 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2807 extensions, and comparison tests, and to fix some bugs.
2809 14. Changes to the "personal" filter test:
2811 (1) The test was buggy in that it was just doing the equivalent of
2812 "contains" tests on header lines. For example, if a user's address was
2813 anne@some.where, the "personal" test would incorrectly be true for
2815 To: susanne@some.where
2817 This test is now done by extracting each address from the header in turn,
2818 and checking the entire address. Other tests that are part of "personal"
2819 are now done using regular expressions (for example, to check local parts
2820 of addresses in From: header lines).
2822 (2) The list of non-personal local parts in From: addresses has been
2823 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2824 taken from the Sieve specification recommendations.
2826 (3) If the message contains any header line starting with "List-" it is
2827 treated as non-personal.
2829 (4) The test for "circular" in the Subject: header line has been removed
2830 because it now seems ill-conceived.
2832 15. Minor typos in src/EDITME comments corrected.
2834 16. Installed latest exipick from John Jetmore.
2836 17. If headers_add on a router specified a text string that was too long for
2837 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2838 of string_sprintf() is now avoided.
2840 18. $message_body_size was not set (it was always zero) when running the DATA
2841 ACL and the local_scan() function.
2843 19. For the "mail" command in an Exim filter, no default was being set for
2844 the once_repeat time, causing a random time value to be used if "once" was
2845 specified. (If the value happened to be <= 0, no repeat happened.) The
2846 default is now 0s, meaning "never repeat". The "vacation" command was OK
2847 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2848 (I found it when inspecting the code).
2850 20. There is now an overall timeout for performing a callout verification. It
2851 defaults to 4 times the callout timeout, which applies to individual SMTP
2852 commands during the callout. The overall timeout applies when there is more
2853 than one host that can be tried. The timeout is checked before trying the
2854 next host. This prevents very long delays if there are a large number of
2855 hosts and all are timing out (e.g. when the network connections are timing
2856 out). The value of the overall timeout can be changed by specifying an
2857 additional sub-option for "callout", called "maxwait". For example:
2859 verify = sender/callout=5s,maxwait=20s
2861 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2862 to the end before writing, but this should make it even safer).
2864 22. Exim was forgetting that it had advertised PIPELINING for the second and
2865 subsequent messages on an SMTP connection. It was also not resetting its
2866 memory on STARTTLS and an internal HELO.
2868 23. When Exim logs an SMTP synchronization error within a session, it now
2869 records whether PIPELINING has been advertised or not.
2871 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2872 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2873 rather than long int.
2875 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2877 26. Added the never_mail option to autoreply.
2883 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2884 crash if the getsockname() call failed; this can happen if a connection is
2885 closed very soon after it is established. The problem was simply in the
2886 order in which certain operations were done, causing Exim to try to write
2887 to the SMTP stream before it had set up the file descriptor. The bug has
2888 been fixed by making things happen in the correct order.
2894 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2895 before the connection was closed, thus losing the rejection response.
2897 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2898 some early Solaris releases, but causes trouble in current releases where
2899 socklen_t is defined.
2901 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2904 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2907 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2908 that releasing the top of it at the end releases what was used for sub-
2909 expansions (unless the block got too big). However, discard this block if
2910 the first thing is a variable or header, so that we can use its block when
2911 it is dynamic (useful for very large $message_headers, for example).
2913 6. Lookups now cache *every* query, not just the most recent. A new, separate
2914 store pool is used for this. It can be recovered when all lookup caches are
2915 flushed. Lookups now release memory at the end of their result strings.
2916 This has involved some general refactoring of the lookup sources.
2918 7. Some code has been added to the store_xxx() functions to reduce the amount
2919 of flapping under certain conditions.
2921 8. log_incoming_interface used to affect only the <= reception log lines. Now
2922 it causes the local interface and port to be added to several more SMTP log
2923 lines, for example "SMTP connection from", and rejection lines.
2925 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2927 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2929 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2930 could be overwritten at the end of the current message (or the start of a
2931 new message if it was set in a HELO ACL). The value is now preserved for
2932 the duration of the SMTP connection.
2934 12. If a transport had a headers_rewrite setting, and a matching header line
2935 contained an unqualified address, that address was qualified, even if it
2936 did not match any rewriting rules. The underlying bug was that the values
2937 of the flags that permit the existence of unqualified sender and recipient
2938 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2939 non-local messages, and by -bnq for local messages) were not being
2940 preserved with the message after it was received.
2942 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2943 "next input=" as part of the text comprising the host identity instead of
2944 the correct text. The code was using the same buffer for two different
2945 strings. However, depending on which order the printing function evaluated
2946 its arguments, the bug did not always show up. Under Linux, for example, my
2947 test suite worked just fine.
2949 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2950 doesn't work with some older versions of Perl. It has been changed to "my",
2951 which in any case is probably the better facility to use.
2953 15. A really picky compiler found some instances of statements for creating
2954 error messages that either had too many or two few arguments for the format
2957 16. The size of the buffer for calls to the DNS resolver has been increased
2958 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2959 for addresses that have a lot of PTR records. This alleviates a problem; it
2960 does not fully solve it.
2962 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2963 buffer now truncates the list and logs the incident, which is the same
2964 action as happens when Exim is looking up a host name and its aliases.
2965 Previously in this situation something unpredictable would happen;
2966 sometimes it was "internal error: store_reset failed".
2968 18. If a server dropped the connection unexpectedly when an Exim client was
2969 using GnuTLS and trying to read a response, the client delivery process
2970 crashed while trying to generate an error log message.
2972 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2973 string, for example:
2975 warn message = H1: something\nH2: something
2977 the text was added as a single header line from Exim's point of view
2978 though it ended up OK in the delivered message. However, searching for the
2979 second and subsequent header lines using $h_h2: did not work. This has been
2980 fixed. Similarly, if a system filter added multiple headers in this way,
2981 the routers could not see them.
2983 20. Expanded the error message when iplsearch is called with an invalid key to
2984 suggest using net-iplsearch in a host list.
2986 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2987 ACLs are no longer actually imposed (and a message to that effect is
2990 22. If a "gecos" field in a passwd entry contained escaped characters, in
2991 particular, if it contained a \" sequence, Exim got it wrong when building
2992 a From: or a Sender: header from that name. A second bug also caused
2993 incorrect handling when an unquoted " was present following a character
2994 that needed quoting.
2996 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2997 was not being matched caselessly.
2999 24. Arranged for all hyphens in the exim.8 source to be escaped with
3002 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3003 properly. Recipient callout cache records were still being keyed to include
3004 the sender, even when use_sender was set false. This led to far more
3005 callouts that were necessary. The sender is no longer included in the key
3006 when use_sender is false.
3008 26. Added "control = submission" modifier to ACLs.
3010 27. Added the ${base62d: operator to decode base 62 numbers.
3012 28. dnsdb lookups can now access SRV records.
3014 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3015 the configuration file.
3017 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3018 (-v) mode. This makes the output for a verbose queue run more intelligible.
3020 31. Added a use_postmaster feature to recipient callouts.
3022 32. Added the $body_zerocount variable, containing the number of binary zero
3023 bytes in the message body.
3025 33. The time of last modification of the "new" subdirectory is now used as the
3026 "mailbox time last read" when there is a quota error for a maildir
3029 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3031 35. Added +ignore_unknown as a special item in host lists.
3033 36. Code for decoding IPv6 addresses in host lists is now included, even if
3034 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3035 address was recognized as an IP address, but was then not correctly decoded
3036 into binary, causing unexpected and incorrect effects when compared with
3043 1. Very minor rewording of debugging text in manualroute to say "list of
3044 hosts" instead of "hostlist".
3046 2. If verify=header_syntax was set, and a header line with an unqualified
3047 address (no domain) and a large number of spaces between the end of the
3048 name and the colon was received, the reception process suffered a buffer
3049 overflow, and (when I tested it) crashed. This was caused by some obsolete
3050 code that should have been removed. The fix is to remove it!
3052 3. When running in the test harness, delay a bit after writing a bounce
3053 message to get a bit more predictability in the log output.
3055 4. Added a call to search_tidyup() just before forking a reception process. In
3056 theory, someone could use a lookup in the expansion of smtp_accept_max_
3057 per_host which, without the tidyup, could leave open a database connection.
3059 5. Added the variables $recipient_data and $sender_data which get set from a
3060 lookup success in an ACL "recipients" or "senders" condition, or a router
3061 "senders" option, similar to $domain_data and $local_part_data.
3063 6. Moved the writing of debug_print from before to after the "senders" test
3066 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3067 problems for message scanning, either using a data ACL, or using
3068 local_scan() because the Received: header was not generated till after they
3069 were called (in order to set the time as the time of reception completion).
3070 I have revised the way this works. The header is now generated after the
3071 body is received, but before the ACL or local_scan() are called. After they
3072 are run, the timestamp in the header is updated.
3078 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3079 before starting a queue runner without re-exec. This happened only when
3080 deliver_drop_privilege was set or when the Exim user was set to root. The
3081 effect of the bug was that timeouts during subsequent deliveries caused
3082 crashes instead of being properly handled. The handler is now left at its
3083 default (and expected) setting.
3085 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3086 message, again when deliver_drop_privilege is set or Exim is run as root.
3087 The bug described in (1) was not present in this case, but the tidying up
3088 of the other signals was missing. I have made the two cases consistent.
3090 3. The ignore_target_hosts setting on a manualroute router was being ignored
3091 for hosts that were looked up using the /MX notation.
3093 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3096 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3097 operated on the sender address. After changing the $sender_address to <>
3098 for the sender address verify, Exim was re-instated it as the original
3099 (before rewriting) address, but remembering that it had rewritten it, so it
3100 wasn't rewriting it again. This bug also had the effect of breaking the
3101 sender address verification caching when the sender address was rewritten.
3103 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3104 This has been changed so that if the ip literal address matches
3105 ignore_target_hosts, the router declines.
3107 7. Added expansion conditions match_domain, match_address, and match_local_
3108 part (NOT match_host).
3110 8. The placeholder for the Received: header didn't have a length field set.
3112 9. Added code to Exim itself and to exim_lock to test for a specific race
3113 condition that could lead to file corruption when using MBX delivery. The
3114 issue is with the lockfile that is created in /tmp. If this file is removed
3115 after a process has opened it but before that process has acquired a lock,
3116 there is the potential for a second process to recreate the file and also
3117 acquire a lock. This could lead to two Exim processes writing to the file
3118 at the same time. The added code performs the same test as UW imapd; it
3119 checks after acquiring the lock that its file descriptor still refers to
3120 the same named file.
3122 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3123 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3126 11. Added the smtp_active_hostname option. If used, this will typically be made
3127 to depend on the incoming interface address. Because $interface_address is
3128 not set up until the daemon has forked a reception process, error responses
3129 that can happen earlier (such as "too many connections") no longer contain
3132 12. If an expansion in a condition on a "warn" statement fails because a lookup
3133 defers, the "warn" statement is abandoned, and the next ACL statement is
3134 processed. Previously this caused the whole ACL to be aborted.
3136 13. Added the iplsearch lookup type.
3138 14. Added ident_timeout as a log selector.
3140 15. Added tls_certificate_verified as a log selector.
3142 16. Added a global option tls_require_ciphers (compare the smtp transport
3143 option of the same name). This controls incoming TLS connections.
3145 17. I finally figured out how to make tls_require_ciphers do a similar thing
3146 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3147 before starting the TLS session.
3149 18. Tabs are now shown as \t in -bP output.
3151 19. If the log selector return_path_on_delivery was set, Exim crashed when
3152 bouncing a message because it had too many Received: header lines.
3154 20. If two routers both had headers_remove settings, and the first one included
3155 a superfluous trailing colon, the final name in the first list and the
3156 first name in the second list were incorrectly joined into one item (with a
3157 colon in the middle).
3163 1. Added -C and -D options to the exinext utility, mainly to make it easier
3164 to include in the automated testing, but these could be helpful when
3165 multiple configurations are in use.
3167 2. The exinext utility was not formatting the output nicely when there was
3168 an alternate port involved in the retry record key, nor when there was a
3169 message id as well (for retries that were specific to a specific message
3170 and a specific host). It was also confused by IPv6 addresses, because of
3171 the additional colons they contain. I have fixed the IPv4 problem, and
3172 patched it up to do a reasonable job for IPv6.
3174 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3175 delivery, the log line now contains "pipelined" if PIPELINING was used.
3177 4. An SMTP transport process used to panic and die if the bind() call to set
3178 an explicit outgoing interface failed. This has been changed; it is now
3179 treated in the same way as a connect() failure.
3181 5. A reference to $sender_host_name in the part of a conditional expansion
3182 that was being skipped was still causing a DNS lookup. This no longer
3185 6. The def: expansion condition was not recognizing references to header lines
3186 that used bh_ and bheader_.
3188 7. Added the _cache feature to named lists.
3190 8. The code for checking quota_filecount in the appendfile transport was
3191 allowing one more file than it should have been.
3193 9. For compatibility with Sendmail, the command line option
3201 and sets the incoming protocol and host name (for trusted callers). The
3202 host name and its colon can be omitted when only the protocol is to be set.
3203 Note the Exim already has two private options, -pd and -ps, that refer to
3204 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3205 "s", but I don't think that's a major issue.
3207 10. A number of refactoring changes to the code, none of which should affect
3210 (a) The number of logging options was getting close to filling up the
3211 32-bit word that was used as a bit map. I have split them into two classes:
3212 those that are passed in the argument to log_write(), and those that are
3213 only ever tested independently outside of that function. These are now in
3214 separate 32-bit words, so there is plenty of room for expansion again.
3215 There is no change in the user interface or the logging behaviour.
3217 (b) When building, for example, log lines, the code previously used a
3218 macro that called string_cat() twice, in order to add two strings. This is
3219 not really sufficiently general. Furthermore, there was one instance where
3220 it was actually wrong because one of the argument was used twice, and in
3221 one call a function was used. (As it happened, calling the function twice
3222 did not affect the overall behaviour.) The macro has been replaced by a
3223 function that can join an arbitrary number of extra strings onto a growing
3226 (c) The code for expansion conditions now uses a table and a binary chop
3227 instead of a serial search (which was left over from when there were very
3228 few conditions). Also, it now recognizes conditions like "pam" even when
3229 the relevant support is not compiled in: a suitably worded error message is
3230 given if an attempt is made to use such a condition.
3232 11. Added ${time_interval:xxxxx}.
3234 12. A bug was causing one of the ddress fields not to be passed back correctly
3235 from remote delivery subprocesses. The field in question was not being
3236 subsequently used, so this caused to problems in practice.
3238 13. Added new log selectors queue_time and deliver_time.
3240 14. Might have fixed a bug in maildirsizefile handling that threw up
3241 "unexpected character" debug warnings, and recalculated the data
3242 unnecessarily. In any case, I expanded the warning message to give more
3245 15. Added the message "Restricted characters in address" to the statements in
3246 the default ACL that block characters like @ and % in local parts.
3248 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3249 Three changes have been made:
3251 (a) There was a serious bug; a negative response to MAIL caused the whole
3252 recipient domain to be cached as invalid, thereby blocking all messages
3253 to all local parts at the same domain, from all senders. This bug has
3254 been fixed. The domain is no longer cached after a negative response to
3255 MAIL if the sender used is not empty.
3257 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3260 (c) A new callout option, "use_sender" has been added for people who want
3261 the modified behaviour.
3267 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3270 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3271 indeed breaks things for older releases.
3273 3. Added additional logging to the case where there is a problem reading data
3274 from a filter that is running in a subprocess using a pipe, in order to
3275 try to track down a specific problem.
3277 4. Testing facility fudge: when running in the test harness and attempting
3278 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3279 getting "No route to host". Convert this to a timeout.
3281 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3284 6. Some OS don't have socklen_t but use size_t instead. This affects the
3285 fifth argument of getsockopt() amongst other things. This is now
3286 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3287 can be set for individual OS. I have set it for SunOS5, OSF1, and
3288 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3289 some earlier ones do not.
3291 7. Change 4.30/15 was not doing the test caselessly.
3293 8. The standard form for an IPv6 address literal was being rejected by address
3294 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3295 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3296 this, as well as the form without the "IPv6" on the front (but only when
3297 address literals are enabled, of course).
3299 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3301 10. Exim crashed if a message with an empty sender address specified by -f
3302 encountered a router with an errors_to setting. This could be provoked only
3303 by a command such as
3307 where an empty string was supplied; "<>" did not hit this bug.
3309 11. Installed PCRE release 4.5.
3311 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3312 remained set. It is now erased.
3314 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3315 times from message ids (which are base 36 rather than the normal 62).
3317 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3318 were being counted as actual protocol errors, and logged if the log
3319 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3320 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3321 and DATA following a set of rejected RCPTs do not count as protocol errors.
3322 In other words, Exim assumes they were pipelined, though this may not
3323 actually be the case. Of course, in all cases the client gets an
3324 appropriate error code.
3326 15. If a lookup fails in an ACL condition, a message about the failure may
3327 be available; it is used if testing the ACL cannot continue, because most
3328 such messages specify what the cause of the deferral is. However, some
3329 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3330 that caused an old message to be retained and used if a later statement
3331 caused a defer, replacing the real cause of the deferral.
3333 16. If an IP address had so many PTR records that the DNS lookup buffer
3334 was not large enough to hold them, Exim could crash while trying to process
3335 the truncated data. It now detects and logs this case.
3337 17. Further to 4.21/58, another change has been made: if (and only if) the
3338 first line of a message (the first header line) ends with CRLF, a bare LF
3339 in a subsequent header line has a space inserted after it, so as not to
3340 terminate the header.
3342 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3343 unnecessarily, used atoi() instead of strtol(), and didn't check the
3344 termination when getting file sizes from file names by regex.
3346 19. Completely re-implemented the support for maildirsize files, in the light
3347 of a number of problems with the previous contributed implementation
3348 (4.30/29). In particular:
3350 . If the quota is zero, the maildirsize file is maintained, but no quota is
3353 . If the maildir directory does not exist, it is created before any attempt
3354 to write a maildirsize file.
3356 . The quota value in the file is just a cache; if the quota is changed in
3357 the transport, the new value overrides.
3359 . A regular expression is available for excluding directories from the
3362 20. The autoreply transport checks the characters in options that define the
3363 message's headers; it allows continued headers, but it was checking with
3364 isspace() after an embedded newline instead of explicitly looking for a
3367 21. If all the "regular" hosts to which an address was routed had passed their
3368 expiry times, and had not reached their retry times, the address was
3369 bounced, even if fallback hosts were defined. Now Exim should go on to try
3372 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3373 equivalent code in the SMTP transport. Some hosts send humungous responses
3374 to HELO/EHLO, more than 1024 it seems.
3376 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3377 gives compiler warnings in some environments. I've now done it "properly",
3380 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3381 (because of problems with the built-in one) was declared to return uschar *
3382 instead of char *, causing compiler failure.
3384 25. Fixed a file descriptor leak when processing alias/forward files.
3386 26. Fixed a minor format string issue in dbfn.c.
3388 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3390 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3391 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3394 29. When a sender address is verified, it is cached, to save repeating the test
3395 when there is more than one recipient in a message. However, when the
3396 verification involves a callout, it is possible for different callout
3397 options to be set for different recipients. It is too complicated to keep
3398 track of this in the cache, so now Exim always runs a verification when a
3399 callout is required, relying on the callout cache for the optimization.
3400 The overhead is duplication of the address routing, but this should not be
3403 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3404 to be verified with callout=postmaster, and the main callout worked but the
3405 postmaster check failed, the verification correctly failed. However, if a
3406 subsequent RCPT command asked for sender verification *without* the
3407 postmaster check, incorrect caching caused this verification also to fail,
3410 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3411 it was not caching the DNS options (qualify_single, search_parents) that
3412 were used when the lookup failed. A subsequent lookup with different
3413 options therefore always gave the same answer, though there were cases
3414 where it should not have. (Example: a "domains = !$mx_any" option on a
3415 dnslookup router: the "domains" option is always processed without any
3416 widening, but the router might have qualify_single set.) Now Exim uses the
3417 cached value only when the same options are set.
3419 32. Added John Jetmore's "exipick" utility to the distribution.
3421 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3422 than a timeout (e.g. a certificate is required, and is not provided), an
3423 Exim server now closes the connection immediately. Previously it waited for
3424 the client to close - but if the client is SSL, it seems that they each
3425 wait for each other, leading to a delay before one of them times out.
3427 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3428 maintained 0.8.x compatibility because I don't think many are using it, and
3429 it is clearly obsolete.
3431 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3434 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3435 client certificate was expired. A simple patch fixes this, though I don't
3436 understand the full logic of why the verify callback is called multiple
3439 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3440 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3441 which causes problems with some clients (such as the Certicom SSL Plus
3442 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3443 disables the coutermeasure allowing Eudora to connect."
3445 38. Exim was not checking that a write() to a log file succeeded. This could
3446 lead to Bad Things if a log got too big, in particular if it hit a file
3447 size limit. Exim now panics and dies if it cannot write to a log file, just
3448 as it does if it cannot open a log file.
3450 39. Modified OS/Makefile-Linux so that it now contains
3452 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3454 The two -D definitions ensure that Exim is compiled with large file
3455 support, which makes it possible to handle log files that are bigger than
3458 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3459 instance) a domain was checked against a named list that involved a lookup,
3460 causing $domain_data to be set, then another domain was checked against the
3461 same list, then the first domain was re-checked, the value of $domain_data
3462 after the final check could be wrong. In particular, if the second check
3463 failed, it could be set empty. This bug probably also applied to
3466 41. The strip_trailing_dot option was not being applied to the address given
3467 with the -f command-line option.
3469 42. The code for reading a message's header from the spool was incrementing
3470 $received_count, but never initializing it. This meant that the value was
3471 incorrect (doubled) while delivering a message in the same process in which
3472 it was received. In the most common configuration of Exim, this never
3473 happens - a fresh exec is done - but it can happen when
3474 deliver_drop_privilege is set.
3476 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3477 it now includes up to 150 characters of the unexpected data in the log
3480 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3481 and building data strings. The size of both of these buffers was 10 000
3482 bytes - far larger than anybody would *ever* want, thought I. Needless to
3483 say, somebody hit the limit. I have increased the maximum line length to
3484 20 000 and the maximum data length of concatenated lines to 100 000. I have
3485 also fixed two bugs, because there was no checking on these buffers. Tsk,
3486 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3487 buffer is too small.
3489 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3490 lsearch lookups. Now it does.
3492 46. When parsing a route_list item in a manualroute router, a fixed-length
3493 buffer was used for the list of hosts. I made this 1024 bytes long,
3494 thinking that nobody would ever have a list of hosts that long. Wrong.
3495 Somebody had a whole pile of complicated expansion conditions, and the
3496 string was silently truncated, leading to an expansion error. It turns out
3497 that it is easier to change to an unlimited length (owing to other changes
3498 that have happened since this code was originally written) than to build
3499 structure for giving a limitation error. The length of the item that
3500 expands into the list of hosts is now unlimited.
3502 47. The lsearch lookup could not handle data where the length of text line was
3503 more than 4095 characters. Such lines were truncated, leading to shortened
3504 data being returned. It should now handle lines of any length.
3506 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3507 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3510 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3511 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3512 no longer happens in comment lines. A list of replacements is now placed
3513 at the head of all of the source files, except those whose only change is
3514 to replace PERL_COMMAND in the very first #! line.
3516 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3517 messages on the queue, with a bottom-up merge sort, using code contributed
3518 by Michael Haardt. This should make operations like -bp somewhat faster on
3519 large queues. It won't affect queue runners, except when queue_run_in_order
3522 51. Installed eximstats 1.31 in the distribution.
3524 52. Added support for SRV lookups to the dnslookup router.
3526 53. If an ACL referred to $message_body or $message_body_end, the value was not
3527 reset for any messages that followed in the same SMTP session.
3529 54. The store-handling optimization for building very long strings was not
3530 differentiating between the different store pools. I don't think this
3531 actually made any difference in practice, but I've tidied it.
3533 55. While running the routers to verify a sender address, $sender_address
3534 was still set to the sender address. This is wrong, because when routing to
3535 send a bounce to the sender, it would be empty. Therefore, I have changed
3536 it so that, while verifying a sender address, $sender_address is set to <>.
3537 (There is no change to what happens when verifying a recipient address.)
3539 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3540 target A or AAAA records (if not already returned) without resetting the
3541 qualify_single or search_parents options of the DNS resolver. These are
3542 inappropriate in this case because the targets of MX and SRV records must
3543 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3544 target that, when qualified, matched something in the local domain. These
3545 two options are now turned off when doing these lookups.
3547 57. It seems that at least some releases of Reiserfs (which does not have the
3548 concept of a fixed number of inodes) returns zero and not -1 for the
3549 number of available inodes. This interacted badly with check_spool_inodes,
3550 which assumed that -1 was the "no such thing" setting. What I have done is
3551 to check that the total number of inodes is greater than zero before doing
3552 the test of how many are available.
3554 58. When a "warn" ACL statement has a log_message modifier, the message is
3555 remembered, and not repeated. This is to avoid a lot of repetition when a
3556 message has many recipients that cause the same warning to be written.
3557 Howewer, Exim was preserving the list of already written lines for an
3558 entire SMTP session, which doesn't seem right. The memory is now reset if a
3559 new message is started.
3561 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3562 debugging output unless log_rewrite was also set.
3564 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3565 of (int)(handle) when we know that handle contains (void *)(-1).
3567 61. The Exim daemon panic-logs an error return when it closes the incoming
3568 connection. However "connection reset by peer" seems to be common, and
3569 isn't really an error worthy of noting specially, so that particular error
3572 62. When Exim is trying to find all the local interfaces, it used to panic and
3573 die if the ioctl to get the interface flags failed. However, it seems that
3574 on at least one OS (Solaris 9) it is possible to have an interface that is
3575 included in the list of interfaces, but for which you get a failure error
3576 for this call. This happens when the interface is not "plumbed" into a
3577 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3578 failure of the "get flags" call assumes that the interface is down.
3580 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3581 makes life easier for people who are doing arithmetic on fields extracted
3582 from dates, where you often get leading zeros that should not be
3583 interpreted as octal.
3585 64. Added qualify_domain to the redirect router, to override the global
3588 65. If a pathologically long header line contained very many addresses (the
3589 report of this problem mentioned 10 000) and each of them was rewritten,
3590 Exim could use up a very large amount of memory. (It kept on making new
3591 copies of the header line as it rewrote, and never released the old ones.)
3592 At the expense of a bit more processing, the header rewriting function has
3593 been changed so that it no longer eats memory in this way.
3595 66. The generation of the Received: header has been moved from the time that a
3596 message starts to be received, to the time that it finishes. The timestamp
3597 in the Received: header should now be very close to that of the <= log
3598 line. There are two side-effects of this change:
3600 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3601 logged header lines no longer include the local Received: line, because
3602 it has not yet been created. The same applies to a copy of the message
3603 that is returned to a non-SMTP sender when a message is rejected.
3605 (b) When a filter file is tested using -bf, no additional Received: header
3606 is added to the test message. After some thought, I decided that this
3609 This change does not affect the value of $received_for. It is still set
3610 after address rewriting, but before local_scan() is called.
3612 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3614 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3615 gave an unhelpful panic error message, and a defer error. I have managed to
3616 change this behaviour so that it now rejects any supplied certificate,
3617 which seems right, as the list of acceptable certificates is empty.
3619 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3620 gave an unhelpful defer error. I have not managed to make this reject any
3621 supplied certificates, but the error message it gives is "no certificate
3622 supplied", which is not helpful.
3624 70. exigrep's output now also includes lines that are not associated with any
3625 message, but which match the given pattern. Implemented by a patch from
3626 Martin Sluka, which also tidied up the Perl a bit.
3628 71. Recipient callout verification, like sender verification, was using <> in
3629 the MAIL FROM command. This isn't really the right thing, since the actual
3630 sender may affect whether the remote host accepts the recipient or not. I
3631 have changed it to use the actual sender in the callout; this means that
3632 the cache record is now keyed on a recipient/sender pair, not just the
3633 recipient address. There doesn't seem to be a real danger of callout loops,
3634 since a callout by the remote host to check the sender would use <>.
3635 [SEE ABOVE: changed after hitting problems.]
3637 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3638 temporary errors. However, in the case of such a code being given after
3639 the end of a data transmission (i.e. after ".") Exim was failing to write
3640 a retry record for the message. (Yes, there was some broken host that was
3641 actually sending 8xx at this point.)
3643 73. An unknown lookup type in a host list could cause Exim to panic-die when
3644 the list was checked. (An example that provoked this was putting <; in the
3645 middle of a list instead of at the start.) If this happened during a DATA
3646 ACL check, a -D file could be left lying around. This kind of configuration
3647 error no longer causes Exim to die; instead it causes a defer errror. The
3648 incident is still logged to the main and panic logs.
3650 74. Buglet left over from Exim 3 conversion. The message "too many messages
3651 in one connection" was written to the rejectlog but not the mainlog, except
3652 when address rewriting (yes!) was being logged.
3654 75. Added write_rejectlog option.
3656 76. When a system filter was run not as root (that is, when system_filter_user
3657 was set), the values of the $n variables were not being returned to the
3658 main process; thus, they were not subsequently available in the $sn
3661 77. Added +return_path_on_delivery log selector.
3663 78. A connection timeout was being treated differently from recipients deferred
3664 when testing hosts_max_try with a message that was older than the host's
3665 retry timeout. (The host should not be counted, thus allowing all hosts to
3666 be tried at least once before bouncing.) This may have been the cause of an
3667 occasionally reported bug whereby a message would remain on the queue
3668 longer than the retry timeout, but would be bounced if a delivery was
3669 forced. I say "may" because I never totally pinned down the problem;
3670 setting up timeout/retry tests is difficult. See also the next item.
3672 79. The ultimate address timeout was not being applied to errors that involved
3673 a combination of host plus message (for example, a timeout on a MAIL
3674 command). When an address resolved to a number of possible hosts, and they
3675 were not all tried for each delivery (e.g. because of hosts_max_try), a
3676 message could remain on the queue longer than the retry timeout.
3678 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3681 81. Fixed an obscure SMTP outgoing bug which required at least the following
3682 conditions: (a) there was another message waiting for the same server;
3683 (b) the server returned 5xx to all RCPT commands in the first message so
3684 that the message was not completed; (c) the server dropped the connection
3685 or gave a negative response to the RSET that Exim sends to abort the
3686 transaction. The observed case was a dropped connection after DATA that had
3687 been sent in pipelining mode. That is, the server had advertised PIPELINING
3688 but was not implementing it correctly. The effect of the bug was incorrect
3689 behaviour, such as trying another host, and this could lead to a crash.
3695 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3696 and daemon.c were passed as pointers to ints; they should have been
3697 pointers to socklen_t variables (which are typically unsigned ints).
3699 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3702 3. Fixed a really odd bug that affected only the testing scheme; patching a
3703 certain fixed string in the binary changed the value of another string that
3704 happened to be identical to the end of the original first string.
3706 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3707 name", it returns that address as the IP address. On some operating
3708 systems (e.g. Solaris), it also passes back the IP address string as the
3709 "host name". However, on others (e.g. Linux), it passes back an empty
3710 string. Exim wasn't checking for this, and was changing the host name to an
3711 empty string, assuming it had been canonicized.
3713 5. Although rare, it is permitted to have more than one PTR record for a given
3714 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3715 all the names associated with an address, because they do in Solaris.
3716 However, it seems that they do not in Linux for data that comes from the
3717 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3718 I found this out when I moved to a new Linux workstation and tried to run
3719 the Exim test suite.
3721 To get round this problem I have changed the code so that it now does its
3722 own call to the DNS to look up PTR records when searching for a host name.
3723 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3724 addresses that are only in /etc/hosts are still found.
3726 This behaviour is, however, controlled by an option called host_lookup_
3727 order, which defaults to "bydns:byaddr". If people want to use the other
3728 order, or indeed, just use one or the other means of lookup, they can
3729 specify it in this variable.
3731 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3732 some operating systems, this comes back from gethostbyaddr() as an empty
3733 string, and this is what Exim used to test for. However, it seems that in
3734 other systems, "." is yielded. Exim now tests for this case too.
3736 7. The values of check_spool_space and check_log_space are now held internally
3737 as a number of kilobytes instead of an absolute number of bytes. If a
3738 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3739 kilobyte. This means that much larger values can be stored.
3741 8. Exim monitor: an attempt to get the action menu when not actually pointing
3742 at a message produces an empty menu entitled "No message selected". This
3743 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3744 no entries in it ("Shell widget menu has zero width and/or height"). So I
3745 have added a single, blank menu entry in this case.
3747 9. Added ${quote_local_part.
3749 10. MIME decoding is now applied to the contents of Subject: header lines when
3752 11. Now that a reference to $sender_host_address automatically causes a reverse
3753 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3754 host lookup before query-style lookups in lists that might use this
3755 variable. This has therefore been abolished, and the "net-" prefix is no
3756 longer necessary for query-style lookups.
3758 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3759 have been a typo for LFLAGS, so it has been changed.
3761 13. The install script calls Exim with "-C /dev/null" in order to find the
3762 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3763 to be output. Howeve, since Exim outputs its version number before the
3764 error, it didn't break the script. It just looked ugly. I fixed this by
3765 always allowing "-C /dev/null" if the caller is root.
3767 14. Ignore overlarge ACL variable number when reading spool file - insurance
3768 against a later release with more variables having written the file.
3770 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3771 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3772 this, as well as the form without the "IPv6" on the front.
3774 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3775 OS/Makefile-Darwin file.
3777 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3778 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3780 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3781 variable. However, some OS use ints for inodes. I've added cast to long int
3782 to get rid of the compiler warning.
3784 19. I had forgotten to lock out "/../" in configuration file names when
3785 ALT_CONFIG_PREFIX was set.
3787 20. Routers used for verification do not need to specify transports. However,
3788 if such a router generated a host list, and callout was configured, Exim
3789 crashed, because it could not find a port number from the (non-existent)
3790 transport. It now assumes port 25 in this circumstance.
3792 21. Added the -t option to exigrep.
3794 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3795 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3796 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3797 them is set, LOOKUP_LSEARCH is forced.
3799 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3801 24. Added sender and host information to the "rejected by local_scan()" log
3802 line; previously there was no indication of these.
3804 25. Added .include_if_exists.
3806 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3807 Linux. It turns out that not all file systems support this. Apparently some
3808 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3809 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3810 not supported on the file descriptor, is now ignored when Exim is trying to
3811 sync a directory. This applies only to Linux.
3813 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3815 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3816 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3819 29. Added support for maildirsize files from supplied patch (modified a bit).
3821 30. The use of :fail: followed by an empty string could lead Exim to respond to
3822 sender verification failures with (e.g.):
3824 550 Verification failed for <xxx>
3825 550 Sender verify failed
3827 where the first response line was missing the '-' that indicates it is not
3828 the final line of the response.
3830 31. The loop for finding the name of the user that called Exim had a hardwired
3831 limit of 10; it now uses the value of finduser_retries, which is used for
3832 all other user lookups.
3834 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3837 33. Exim was neglecting to zero errno before one call of strtol() when
3838 expanding a string and expecting an integer value. On some systems this
3839 resulted in spurious "integer overflow" errors. Also, it was casting the
3840 result into an int without checking.
3842 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3843 did not work. The code looks as if it has *never* worked, though it appears
3844 to have been documented since at least releast 1.62. I have made it work.
3846 35. The "timeout_DNS" error in retry rules, also documented since at least
3847 1.62, also never worked. As it isn't clear exactly what this means, and
3848 clearly it isn't a major issue, I have abolished the feature by treating it
3849 as "timeout", and writing a warning to the main and panic logs.
3851 36. The display of retry rules for -brt wasn't always showing the error code
3854 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3855 timeout_connect_A, timeout_connect_MX.
3857 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3858 to the empty sender.
3860 39. The daemon was not analysing the content of -oX till after it had closed
3861 stderr and disconnected from the controlling terminal. This meant that any
3862 syntax errors were only noted on the panic log, and the return code from
3863 the command was 0. By re-arranging the code a little, I've made the
3864 decoding happen first, so such errors now appear on stderr, and the return
3865 code is 1. However, the actual setting up of the sockets still happens in
3866 the disconnected process, so errors there are still only recorded on the
3869 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3870 connections (as happens on some IP stacks) was logged at start up time as
3871 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3872 it from "IPv6 and IPv4", which means that two separate sockets are being
3875 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3876 says whether AF_INET or AF_INET6 was passed as an argument.
3878 42. Exiwhat output was messed up when time zones were included in log
3881 43. Exiwhat now gives more information about the daemon's listening ports,
3882 and whether -tls-on-connect was used.
3884 44. The "port" option of the smtp transport is now expanded.
3886 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3887 silently ignored. Now an error message is written to the main and panic
3890 46. There's a new ACL modifier called "logwrite" which writes to a log file
3891 as soon as it is encountered.
3893 47. Added $local_user_uid and $local_user_gid at routing time.
3895 48. Exim crashed when trying to verify a sender address that was being
3898 49. Exim was recognizing only a space character after ".include". It now also
3899 recognizes a tab character.
3901 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3902 extracting the relevant information from the specification. The man page no
3903 longer contains scrambled data for the -d option, and I've added a section
3904 at the front about calling Exim under different names.
3906 51. Added "extra_headers" argument to the "mail" command in filter files.
3908 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3911 53. Installed eximstats 1.29.
3913 54. Added transport_filter_timeout as a generic transport option.
3915 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3916 Cc: header lines. This was required by RFC 822, but it not required by RFC
3919 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3920 incoming messages that did not have them. Now it does so only if the
3921 message originates locally, that is, if there is no associated remote host
3922 address. When Resent- header lines are present, this applies to the Resent-
3923 lines rather than the non-Resent- lines.
3925 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3926 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3928 58. Messages for configuration errors now include the name of the main
3929 configuration file - useful now that there may be more than one file in a
3930 list (.included file names were always shown).
3932 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3933 for those rare installations that do not start the daemon as root or run it
3934 setuid root. I've cut out the call to initgroups() if the daemon is not
3937 60. The Exim user and group can now be bound into the binary as text strings
3938 that are looked up at the start of Exim's processing.
3940 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3942 62. Added $mailstore_basename variable.
3944 63. Installed patch to sieve.c from Michael Haardt.
3946 64. When Exim failed to open the panic log after failing to open the main log,
3947 the original message it was trying to log was written to stderr and debug
3948 output, but if they were not available (the usual case in production), it
3949 was lost. Now it is written to syslog before the two lines that record the
3950 failures to open the logs.
3952 65. Users' Exim filters run in subprocesses under the user's uid. It is
3953 possible for a "deliver" command or an alias in a "personal" command to
3954 provoke an address rewrite. If logging of address rewriting is configured,
3955 this fails because the process is not running as root or exim. There may be
3956 a better way of dealing with this, but for the moment (because 4.30 needs
3957 to be released), I have disabled address rewrite logging when running a
3958 filter in a non-root, non-exim process.
3964 1. The buildconfig auxiliary program wasn't quoting the value set for
3965 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3966 not defined. This bug was masked in 4.22 by the effect that was fixed in
3969 2. Some messages that were rejected after a message id was allocated were
3970 shown as "incomplete" by exigrep. It no longer does this for messages that
3971 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3973 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3974 have allow_domain_literals set, the ID did not get logged in the <= line.
3975 Domain literals are now always recognized in Message-ID: header lines.
3977 4. The first argument for a ${extract expansion item is the key name or field
3978 number. Leading and trailing spaces in this item were not being ignored,
3979 causing some misleading effects.
3981 5. When deliver_drop_privilege was set, single queue runner processes started
3982 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3983 same command in the process it spins off) were not dropping privilege.
3985 6. When the daemon running as "exim" started a queue runner, it always
3986 re-executed Exim in the spun-off process. This is a waste of effort when
3987 deliver_drop_privilege is set. The new process now just calls the
3988 queue-runner function directly.
3994 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3997 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3998 ignored. Though the use of -oP was forcing the writing of a pid file, it
3999 was always written to the default place.
4001 3. If the message "no IP address found for host xxxx" is generated during
4002 incoming verification, it is now followed by identification of the incoming
4003 connection (so you can more easily find what provoked it).
4005 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4007 5. Added some features to "harden" Exim a bit more against certain attacks:
4009 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4010 be put in Local/Makefile. This is like the never_users runtime option,
4011 but it cannot be overridden. The default setting is "root".
4013 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4014 prefix string with which any file named in a -C command line option
4017 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4018 is retained for -C and -D only if the caller of Exim is root. Without
4019 it, the exim user may also use -C and -D and retain privilege.
4021 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4022 command line option is disabled.
4024 6. Macro names set by the -D option must start with an upper case letter, just
4025 like macro names defined in the configuration file.
4027 7. Added "dereference=" facility to LDAP.
4029 8. Two instances of the typo "uknown" in the source files are fixed.
4031 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4032 the Configure-Makefile script screwed up while processing it.
4034 10. Incorporated PCRE 4.4.
4036 11. The SMTP synchronization check was not operating right at the start of an
4037 SMTP session. For example, it could not catch a HELO sent before the client
4038 waited for the greeting. There is now a check for outstanding input at the
4039 point when the greeting is written. Because of the duplex, asynchronous
4040 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4041 way, but not yet received, when the check is performed.
4043 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4044 on TCP/IP sockets, because this apparently causes some broken clients to
4047 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4048 unchanged) from the Cygwin maintainer.
4050 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4051 was supported instead of testing for mbx. Effectively a typo.
4053 15. The spa authenticator server code was not checking that the input it
4054 received was valid base64.
4056 16. The debug output line for the "set" modifier in ACLs was not showing the
4057 name of the variable that was being set.
4059 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4061 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4062 lookup on its own. Something else was needed to trigger the lookup. For
4063 example, a match in host_lookup or the need for a host name in a host list.
4064 Now, if $sender_host_name is referenced and the host name has not yet been
4065 looked up, a lookup is performed. If the lookup fails, the variable remains
4066 empty, and $host_lookup_failed is set to "1".
4068 19. Added "eqi" as a case-independent comparison operator.
4070 20. The saslauthd authentication condition could segfault if neither service
4071 nor realm was specified.
4073 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4074 error message that was logged was misleading, and incoming SMTP
4075 connections were dropped. The message is now more accurate, and temporary
4076 errors are given to SMTP connections.
4078 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4079 (except RSET) with a 421 error, until QUIT is received. However, it was
4080 failing to send a response to QUIT.
4082 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4083 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4084 ACL is now run after the helo_try_verify_hosts code.
4086 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4087 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4088 case-independent, so other case variants are also recognized). Apparently
4089 some systems use these upper case variants.
4091 25. If more than two messages were waiting for the same host, and a transport
4092 filter was specified for the transport, Exim sent two messages over the
4093 same TCP/IP connection, and then failed with "socket operation on non-
4094 socket" when it tried to send the third.
4096 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4098 27. The extern definition of crypt16() in expand.c was not being excluded when
4099 the OS had its own crypt16() function.
4101 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4102 as a preferred synonym for return_size_limit, both as an option and as an
4105 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4107 30. Changed the default configuration ACL to relax the local part checking rule
4108 for addresses that are not in any local domains. For these addresses,
4109 slashes and pipe symbols are allowed within local parts, but the sequence
4110 /../ is explicitly forbidden.
4112 31. SPA server authentication was not clearing the challenge buffer before
4115 32. log_message in a "warn" ACL statement was writing to the reject log as
4116 well as to the main log, which contradicts the documentation and doesn't
4117 seem right (because no rejection is happening). So I have stopped it.
4119 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4120 However, I am unable to do any testing of this.
4122 34. Fixed an infelicity in the appendfile transport. When checking directories
4123 for a mailbox, to see if any needed to be created, it was accidentally
4124 using path names with one or more superfluous leading slashes; tracing
4125 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4127 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4128 contents are added to the log line that is written for every discarded
4129 recipient. (Previously a log_message setting was ignored.)
4131 36. The ${quote: operator now quotes the string if it is empty.
4133 37. The install script runs exim in order to find its version number. If for
4134 some reason other than non-existence or emptiness, which it checks, it
4135 could not run './exim', it was installing it with an empty version number,
4136 i.e. as "exim-". This error state is now caught, and the installation is
4139 38. An argument was missing from the function that creates an error message
4140 when Exim fails to connect to the socket for saslauthd authentication.
4141 This could cause Exim to crash, or give a corrupted message.
4143 39. Added isip, isip4, and isip6 to ${if conditions.
4145 40. The ACL variables $acl_xx are now saved with the message, and can be
4146 accessed later in routers, transports, and filters.
4148 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4149 strings in the file are not string-expanded.
4151 42. If a MAIL command specified a SIZE value that was too large to fit into an
4152 int variable, the check against message_size_limit failed. Such values are
4153 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4154 day this will have to be increased, but I don't think I want to be around
4155 when emails are that large.
4162 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4163 iconv() is not standard in FreeBSD.
4165 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4166 IPv6 enabled. The observed symptom was a segmentation fault on return from
4167 the function os_common_find_running_interfaces() in src/os.c.
4169 3. In the check_special_case() function in daemon.c I had used "errno" as an
4170 argument name, which causes warnings on some systems. This was basically a
4171 typo, since it was named "eno" in the comments!
4173 4. The code that waits for the clock to tick (at a resolution of some fraction
4174 of a second) so as to ensure message-id uniqueness was always waiting for
4175 at least one whole tick, when it could have waited for less. [This is
4176 almost certainly not relevant at current processor speeds, where it is
4177 unlikely to ever wait at all. But we try to future-proof.]
4179 5. The function that sleeps for a time interval that includes fractions of a
4180 second contained a race. It did not block SIGALRM between setting the
4181 timer, and suspending (a couple of lines later). If the interval was short
4182 and the sigsuspend() was delayed until after it had expired, the suspension
4183 never ended. On busy systems this could lead to processes getting stuck for
4186 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4187 process, before it forks any delivery processes. The open lookup caching
4188 mechanism meant that the open file or database connection was passed into
4189 the delivery process. The problem was that delivery processes always tidy
4190 up cached lookup data. This could cause a problem for the next delivery
4191 process started by the queue runner, because the external queue runner
4192 process does not know about the closure. So the next delivery process
4193 still has data in the lookup cache. In the case of a file lookup, there was
4194 no problem because closing a file descriptor in a subprocess doesn't affect
4195 the parent. However, if the lookup was caching a connection to a database,
4196 the connection was closed, and the second delivery process was likely to
4197 see errors such as "PGSQL: query failed: server closed the connection
4198 unexpectedly". The problem has been fixed by closing all cached lookups
4199 in a queue runner before running a delivery process.
4201 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4202 seem to have the "const" qualifier which it has on other OS. I've
4205 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4206 *of the same type* (client or server) with the same public name that an
4207 error should be diagnosed.
4209 9. When Exim looked up a host name for an IP address, but failed to find the
4210 original IP address when looking up the host name (a safety check), it
4211 output the message "<ip address> does not match any IP for NULL", which was
4212 confusing, to say the least. The bug was that the host name should have
4213 appeared instead of "NULL".
4215 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4216 user that is built into the binary, and the -C or -D options is used, root
4217 privilege is dropped before the configuration file is read. In addition,
4218 logging is switched to stderr instead of the normal log files. If the
4219 configuration then re-defines the Exim user, the unprivileged environment
4220 is probably not what is expected, so Exim logs a panic warning message (but
4223 However, if deliver_drop_privilege is set, the unprivileged state may well
4224 be exactly what is intended, so the warning has been cut out in that case,
4225 and Exim is allowed to try to write to its normal log files.
4231 1. smtp_return_error_details was not giving details for temporary sender
4232 or receiver verification errors.
4234 2. Diagnose a configuration error if two authenticators have the same public
4237 3. Exim used not to create the message log file for a message until the first
4238 delivery attempt. This could be confusing when incoming messages were held
4239 for policy or load reasons. The message log file is now created at the time
4240 the message is received, and an initial "Received" line is written to it.
4242 4. The automatically generated man page for command line options had a minor
4243 bug that caused no ill effects; however, a more serious problem was that
4244 the procedure for building the man page automatically didn't always
4245 operate. Consequently, release 4.20 contains an out-of-date version. This
4246 shouldn't happen again.
4248 5. When building Exim with embedded Perl support, the script that builds the
4249 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4250 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4252 6. The freeze_tell option was not being used for messages that were frozen on
4253 arrival, either by an ACL or by local_scan().
4255 7. Added the smtp_incomplete_transaction log selector.
4257 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4258 was accepting AUTH without a new EHLO.
4260 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4261 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4264 10. Logging of TCP/IP connections (when configured) now happens in the main
4265 daemon process instead of the child process, so that the TCP/IP connection
4266 count is more accurate (but it can never be perfect).
4268 11. The use of "drop" in a nested ACL was not being handled correctly in the
4269 outer ACL. Now, if condition failure induced by the nested "drop" causes
4270 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4271 or "require"), the connection is dropped.
4273 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4274 that yield "discard" can now be used with an "accept" or a "discard" verb,
4275 but an error is generated for any others (because I can't see a useful way
4276 to define what should happen).
4278 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4279 are now processed in the same way as lines in the Exim configuration file.
4280 In particular, continuation lines are supported.
4282 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4284 15. Added -ti meaning -t -i.
4286 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4287 domains, and warn by logging if others are found.
4289 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4290 returned from GIFCONF when Exim is trying to find the list of interfaces on
4291 a host. The code in os.c has been modified to copy each ifreq to an aligned
4292 structure in all cases.
4294 Also, in some cases, the returned ifreq's were being copied to a 'struct
4295 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4296 means the last couple of bytes of an IPv6 address could be chopped if the
4297 ifreq contained only a normal sockaddr (14 bytes storage).
4299 18. Named domain lists were not supported in the hosts_treat_as_local option.
4300 An entry such as +xxxx was not recognized, and was treated as a literal
4303 19. Ensure that header lines added by a DATA ACL are included in the reject log
4304 if the ACL subsequently rejects the message.
4306 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4307 MD5 (which is deprecated).
4309 21. When testing a filter file using -bf, Exim was writing a message when it
4310 took the sender from a "From " line in the message, but it was not doing so
4311 when it took $return_path from a Return-Path: header line. It now does.
4313 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4314 with a valid header line field name (a series of printing characters
4315 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4317 23. Changed "disc" in the source to "disk" to conform to the documentation and
4318 the book and for uniformity.
4320 24. Ignore Sendmail's -Ooption=value command line item.
4322 25. When execve() failed while trying to run a command in a pipe transport,
4323 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4324 could be confused with a return value of 69 from the command itself. This
4325 has been changed to 127, the value the shell returns if it is asked to run
4326 a non-existent command. The wording for the related log line suggests a
4327 non-existent command as the problem.
4329 26. If received_header_text expands to an empty string, do not add a Received:
4330 header line to the message. (Well, it adds a token one on the spool, but
4331 marks it "old" so that it doesn't get used or transmitted.)
4333 27. Installed eximstats 1.28 (addition of -nt option).
4335 28. There was no check for failure on the call to getsockname() in the daemon
4336 code. This can fail if there is a shortage of resources on the system, with
4337 ENOMEM, for example. A temporary error is now given on failure.
4339 29. Contrary to the C standard, it seems that in some environments, the
4340 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4341 program. Exim now does this explicitly; it affects the formatting of
4342 timestamps using strftime().
4344 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4345 complaints. I've now initialized all the variables, to avoid this.
4347 32. Header lines added by a system filter were not being "seen" during
4348 transport-time rewrites.
4350 33. The info_callback() function passed to OpenSSL is set up with type void
4351 (*)(SSL *, int, int), as described somewhere. However, when calling the
4352 function (actually a macro) that sets it up, the type void(*)() is
4353 expected. I've put in a cast to prevent warnings from picky compilers.
4355 34. If a DNS black list lookup found a CNAME record, but there were no A
4356 records associated with the domain it pointed at, Exim crashed.
4358 35. If a DNS black list lookup returned more than one A record, Exim ignored
4359 all but the first. It now scans all returned addresses if a particular IP
4360 value is being sought. In this situation, the contents of the
4361 $dnslist_value variable are a list of all the addresses, separated by a
4364 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4365 to accept a host name and all its aliases if the forward lookup for any of
4366 them yielded the IP address of the incoming connection. Now it accepts only
4367 those names whose forward lookup yields the correct IP address. Any other
4368 names are discarded. This closes a loophole whereby a rogue DNS
4369 administrator could create reverse DNS records to break through a
4370 wildcarded host restriction in an ACL.
4372 37. If a user filter or a system filter that ran in a subprocess used any of
4373 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4374 the wrong values were passed to the pipe command ($thisaddress had the
4375 value of $0, $0 had the value of $1, etc). This bug was introduced by
4376 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4378 38. Improved the line breaking for long SMTP error messages from ACLs.
4379 Previously, if there was no break point between 40 and 75 characters, Exim
4380 left the rest of the message alone. Two changes have been made: (a) I've
4381 reduced the minimum length to 35 characters; (b) if it can't find a break
4382 point between 35 and 75 characters, it looks ahead and uses the first one
4383 that it finds. This may give the occasional overlong line, but at least the
4384 remaining text gets split now.
4386 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4387 file descriptors might be low, and that setting 1000 would always raise it.
4388 It turns out that in some environments, the limit is already over 1000 and
4389 that lowering it causes trouble. So now Exim takes care not to decrease it.
4391 40. When delivering a message, the value of $return_path is set to $sender_
4392 address at the start of routing (routers may change the value). By an
4393 oversight, this default was not being set up when an address was tested by
4394 -bt or -bv, which affected the outcome if any router or filter referred to
4397 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4398 log only when "message" or "log_message" are set. However, if one of the
4399 conditions was an address verification, or a call to a nested ACL, the
4400 messages generated by the underlying test were being passed through. This
4401 no longer happens. The underlying message is available in $acl_verify_
4402 message for both "message" and "log_message" expansions, so it can be
4403 passed through if needed.
4405 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4406 new expansion $bh_ to give the encoded byte string without charset
4407 translation. Translation happens only if iconv() is available; HAVE_ICONV
4408 indicates this at build time. HEADERS_CHARSET gives the charset to
4409 translate to; headers_charset can change it in the configuration, and
4410 "headers charset" can change it in an individual filter file.
4412 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4413 that creates RFC 2047 encoded "words" labels them as that charset instead
4414 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4415 expansion operator; (ii) when Exim creates a From: line for a local
4416 message; (iii) when a header line is rewritten to include a "phrase" part.
4418 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4419 buggy, causing it to skip the first lines of messages whose message ID
4420 ended in 'D'. This would not have bitten before Exim release 4.14, because
4421 message IDs were unlikely to end in 'D' before then. The effect was to have
4422 incorrect size information for certain domains.
4424 45. #include "config.h" was missing at the start of the crypt16.c module. This
4425 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4428 46. If there was a timeout during a "random" callout check, Exim treated it as
4429 a failure of the random address, and carried on sending RSET and the real
4430 address. If the delay was just some slowness somewhere, the response to the
4431 original RCPT would be taken as a response to RSET and so on, causing
4432 mayhem of various kinds.
4434 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4435 when I implemented it. It didn't allow for the fact that some option values
4436 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4437 the right test for positive values.
4439 48. Domain names in DNS records are case-independent. Exim always looks them up
4440 in lower case. Some resolvers return domain names in exactly the case they
4441 appear in the zone file, that is, they may contain uppercase letters. Not
4442 all resolvers do this - some return always lower case. Exim was treating a
4443 change of case by a resolver as a change of domain, similar to a widening
4444 of a domain abbreviation. This triggered its re-routing code and so it was
4445 trying to route what was effectively the same domain again. This normally
4446 caused routing to fail (because the router wouldn't handle the domain
4447 twice). Now Exim checks for this case specially, and just changes the
4448 casing of the domain that it ultimately uses when it transmits the message
4451 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4454 50. If a filter generated a file delivery with a non-absolute name (possible if
4455 no home directory exists for the router), the forbid_file option was not
4458 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4459 the existing equality matching.
4461 52. Exim was using ints instead of ino_t variables in some places where it was
4462 dealing with inode numbers.
4464 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4465 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4466 called TMPDIR, and if it finds it is different, it changes its value.
4468 54. The smtp_printf() function is now made available to local_scan() so
4469 additional output lines can be written before returning. There is also an
4470 smtp_fflush() function to enable the detection of a dropped connection.
4471 The variables smtp_input and smtp_batched_input are exported to
4474 55. Changed the default runtime configuration: the message "Unknown user"
4475 has been removed from the ACL, and instead placed on the localuser router,
4476 using the cannot_route_message feature. This means that any verification
4477 failures that generate their own messages won't get overridden. Similarly,
4478 the "Unrouteable address" message that was in the ACL for unverifiable
4479 relay addresses has also been removed.
4481 56. Added hosts_avoid_esmtp to the smtp transport.
4483 57. The exicyclog script was not checking for the esoteric option
4484 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4485 will work only if exicyclog is run under the appropriate euid.
4487 58. Following a discussion on the list, the rules by which Exim recognises line
4488 endings on incoming messages have been changed. The -dropcr and drop_cr
4489 options are now no-ops, retained only for backwards compatibility. The
4490 following line terminators are recognized: LF CRLF CR. However, special
4491 processing applies to CR:
4493 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4494 nor a local message in the state where . is a terminator.
4496 (ii) If a bare CR is encountered in a header line, an extra space is added
4497 after the line terminator so as not to end the header. The reasoning
4498 behind this is that bare CRs in header lines are most likely either
4499 to be mistakes, or people trying to play silly games.
4501 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4502 was being incorrectly given as 18 bytes larger than it should have been.
4503 This is a VOB (very old bug).
4505 60. This may never have affected anything current, but just in case it has:
4506 When the local host is found other than at the start of a list of hosts,
4507 the local host, those with the same MX, and any that follow, are discarded.
4508 When the list in question was part of a longer list of hosts, the following
4509 hosts (not currently being processed) were also being discarded. This no
4510 longer happens. I'm not sure if this situation could ever has previously
4513 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4516 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4517 header. This is something that is recommended in a new Internet Draft, and
4518 is something that is documented as being done by Sendmail. There are two
4519 possible values. For messages generated by the autoreply transport, Exim
4522 Auto-Submitted: auto-replied
4524 whereas for all other generated messages (e.g. bounces) it adds
4526 Auto-Submitted: auto-generated
4528 63. The "personal" condition in filters now includes a test for the
4529 Auto-Submitted: header. If it contains the string "auto-" the message it
4530 not considered personal.
4532 64. Added rcpt_include_affixes as a generic transport option.
4534 65. Added queue_only_override (default true).
4536 66. Added the syslog_duplication option.
4538 67. If what should have been the first header line of a message consisted of
4539 a space followed by a colon, Exim was mis-interpreting it as a header line.
4540 It isn't of course - it is syntactically invalid and should therefore be
4541 treated as the start of the message body. The misbehaviour could have
4542 caused a number of strange effects, including loss of data in subsequent
4543 header lines, and spool format errors.
4545 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4546 client host had authenticated. This control can now be exercised by an ACL
4547 for more flexibility.
4549 69. By default, callouts do not happen when testing with -bh. There is now a
4550 variant, -bhc, which does actually run the callout code, including
4551 consulting and updating the callout cache.
4553 70. Added support for saslauthd authentication, courtesy of Alexander
4556 71. If statvfs() failed on the spool or log directories while checking their
4557 size for availability, Exim confusingly gave the error "space shortage".
4558 Furthermore, in debugging mode it crashed with a floating point exception.
4559 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4560 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4561 serious problem, Exim now writes to the main and panic logs when this
4562 happens, with details of the failure. It then refuses to accept the
4563 incoming message, giving the message "spool directory problem" or "log
4564 directory problem" with a 421 code for SMTP messages.
4566 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4567 0, 1, and 2 exist, because some OS complain for execs without them (see
4568 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4569 descriptors. However, the code omitted to check that the open succeeded,
4570 causing mysterious errors if for some reason the permissions on /dev/null
4571 got screwed. Now Exim writes a message to the main and panic logs, and
4572 bombs out if it can't open /dev/null.
4574 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4575 interact so that it is all more flexible. It is supposed to remain
4576 backwards compatible. Also added extra_local_interfaces.
4578 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4579 to bomb out with an assertion failure - to the client this appears as a
4580 connection drop. This problem occurs in the part of the code that was taken
4581 from the Samba project. Fortunately, the assertion is in a very simple
4582 function, so I have fixed this by reproducing the function inline in the
4583 one place where it is called, and arranging for authentication to fail
4584 instead of killing the process with assert().
4586 75. The SPA client code was not working when the server requested OEM rather
4587 than Unicode encoding.
4589 76. Added code to make require_files with a specific uid setting more usable in
4590 the case where statting the file as root fails - usually a non-root-mounted
4591 NFS file system. When this happens and the failure is EACCES, Exim now
4592 forks a subprocess and does the per-uid checking as the relevant uid.
4594 77. Added process_log_path.
4596 78. If log_file_path was not explicitly set, a setting of check_log_space or
4597 check_log_inodes was ignored.
4599 79. If a space check for the spool or log partitions fails, the incident is now
4600 logged. Of course, in the latter case the data may get lost...
4602 80. Added the %p formatting code to string_format() so that it can be used to
4603 print addresses in debug_print(). Adjusted all the address printing in the
4604 debugging in store.c to use %p rather than %d.
4606 81. There was a concern that a line of code in smtp_in.c could overflow a
4607 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4608 initially expressed, the concern was not well-founded, because trailing
4609 spaces are removed early. However, if the trailing spaces were followed by
4610 a NULL, they did not get removed, so the overflow was possible. Two fixes
4613 (a) I re-wrote the offending code in a cleaner fashion.
4614 (b) If an incoming SMTP command contains a NULL character, it is rejected
4617 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4618 runs initgroups(), so that if the Exim user is in any additional groups,
4619 they will be used during message reception.
4625 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob