1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.482 2007/02/20 09:53:41 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.
109 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
110 expansion side effects.
112 PH/27 When a message times out after an over-quota error from an Exim-imposed
113 quota, the bounce message says "mailbox is full". This message was not
114 being given when it was a system quota that was exceeded. It now should
117 MH/03 Made $recipients available in local_scan(). local_scan() already has
118 better access to the recipient list through recipients_list[], but
119 $recipients can be useful in postmaster-provided expansion strings.
125 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
126 fixed by 4.65/MH/01 (is this a record?) are fixed:
128 (i) An empty string was always treated as zero by the numeric comparison
129 operators. This behaviour has been restored.
131 (ii) It is documented that the numeric comparison operators always treat
132 their arguments as decimal numbers. This was broken in that numbers
133 starting with 0 were being interpreted as octal.
135 While fixing these problems I realized that there was another issue that
136 hadn't been noticed. Values of message_size_limit (both the global option
137 and the transport option) were treated as octal if they started with 0.
138 The documentation was vague. These values are now always treated as
139 decimal, and I will make that clear in the documentation.
145 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
146 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
149 MH/01 Don't check that the operands of numeric comparison operators are
150 integers when their expansion is in "skipping" mode (fixes bug
151 introduced by 4.64-PH/07).
153 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
154 child addresses, Exim now panics and dies. Previously, because the count
155 is held in a short int, deliveries were likely to be lost. As such a
156 large number of recipients for a single message is ridiculous
157 (performance will be very, very poor), I have chosen to impose a limit
158 rather than extend the field.
164 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
165 leftover -K file (the existence of which was triggered by #402).
166 While we were at it, introduced process PID as part of the -K
167 filename. This should rule out race conditions when creating
170 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
171 processing considerably. Previous code took too long for large mails,
172 triggering a timeout which in turn triggers #401.
174 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
175 in the DK code in transports.c. sendfile() is not really portable,
176 hence the _LINUX specificness.
178 TF/01 In the add_headers option to the mail command in an Exim filter,
179 there was a bug that Exim would claim a syntax error in any
180 header after the first one which had an odd number of characters
183 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
184 callout verification, Exim cached a "reject" for the entire domain. This
185 is correct for most verifications, but it is not correct for a recipient
186 verification with use_sender or use_postmaster set, because in that case
187 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
188 case of MAIL FROM:<> rejection from other early rejections (e.g.
189 rejection of HELO). When verifying a recipient using a non-null MAIL
190 address, the cache is ignored if it shows MAIL FROM:<> rejection.
191 Whatever the result of the callout, the value of the domain cache is
192 left unchanged (for any other kind of callout, getting as far as trying
193 RCPT means that the domain itself is ok).
195 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
198 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
199 manifest itself as EPIPE rather than ECONNECT. When tidying away a
200 session, the daemon ignores ECONNECT errors and logs others; it now
201 ignores EPIPE as well.
203 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
204 (quoted-printable decoding).
206 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
207 later the small subsequent patch to fix an introduced bug.
209 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
211 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
213 PH/08 An error is now given if message_size_limit is specified negative.
215 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
216 to be given (somewhat) arbitrary names.
218 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
221 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
222 miscellaneous code fixes
224 PH/10 Added the log_reject_target ACL modifier to specify where to log
227 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
228 hostname. This is wrong, because it relates to the incoming message (and
229 probably the interface on which it is arriving) and not to the outgoing
230 callout (which could be using a different interface). This has been
231 changed to use the value of the helo_data option from the smtp transport
232 instead - this is what is used when a message is actually being sent. If
233 there is no remote transport (possible with a router that sets up host
234 addresses), $smtp_active_hostname is used.
236 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
237 tweaks were necessary in order to get it to work (see also 21 below):
238 (a) The code assumed that strncpy() returns a negative number on buffer
239 overflow, which isn't the case. Replaced with Exim's string_format()
241 (b) There were several signed/unsigned issues. I just did the minimum
242 hacking in of casts. There is scope for a larger refactoring.
243 (c) The code used strcasecmp() which is not a standard C function.
244 Replaced with Exim's strcmpic() function.
245 (d) The code set only $1; it now sets $auth1 as well.
246 (e) A simple test gave the error "authentication client didn't specify
247 service in request". It would seem that Dovecot has changed its
248 interface. Fortunately there's a specification; I followed it and
249 changed what the client sends and it appears to be working now.
251 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
254 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
255 address A is aliased to B and C, where B exists and C does not. Without
256 -v the output is "A verified" because verification stops after a
257 successful redirection if more than one address is generated. However,
258 with -v the child addresses are also verified. Exim was outputting "A
259 failed to verify" and then showing the successful verification for C,
260 with its parentage. It now outputs "B failed to verify", showing B's
261 parentage before showing the successful verification of C.
263 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
264 look up a TXT record in a specific list after matching in a combined
267 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
268 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
269 they consult the DNS. I had assumed they would set it the way they
270 wanted; and indeed my experiments on Linux seem to show that in some
271 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
272 To be on the safe side, however, I have now made the interface to
273 host_find_byname() similar to host_find_bydns(), with an argument
274 containing the DNS resolver options. The host_find_byname() function now
275 sets these options at its start, just as host_find_bydns() does. The smtp
276 transport options dns_qualify_single and dns_search_parents are passed to
277 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
278 of host_find_byname() use the default settings of RES_DEFNAMES
279 (qualify_single) but not RES_DNSRCH (search_parents).
281 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
282 spool_read_header() do less string testing, by means of a preliminary
283 switch on the second character of optional "-foo" lines. (This is
284 overdue, caused by the large number of possibilities that now exist.
285 Originally there were few.) While I was there, I also converted the
286 str(n)cmp tests so they don't re-test the leading "-" and the first
287 character, in the hope this might squeeze out yet more improvement.
289 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
290 flag allowing group syntax was set by the header_syntax check but not
291 turned off, possible causing trouble later; (2) The flag was not being
292 set at all for the header_verify test, causing "group"-style headers to
293 be rejected. I have now set it in this case, and also caused header_
294 verify to ignore an empty address taken from a group. While doing this, I
295 came across some other cases where the code for allowing group syntax
296 while scanning a header line wasn't quite right (mostly, not resetting
297 the flag correctly in the right place). These bugs could have caused
298 trouble for malformed header lines. I hope it is now all correct.
300 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
301 with the "reply" argument non-NULL. The code, however (which originally
302 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
303 but it didn't always do it. This confused somebody who was copying the
304 code for some other use. I have removed all the tests.
306 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
307 feature that was used to support insecure browsers during the U.S. crypto
308 embargo. It requires special client support, and Exim is probably the
309 only MTA that supported it -- and would never use it because real RSA is
310 always available. This code has been removed, because it had the bad
311 effect of slowing Exim down by computing (never used) parameters for the
312 RSA_EXPORT functionality.
314 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
315 authenticator to fail if there's a tab character in the incoming data
316 (there should never be unless someone is messing about, as it's supposed
317 to be base64-encoded). Also added, on Timo's advice, the "secured" option
318 if the connection is using TLS or if the remote IP is the same as the
319 local IP, and the "valid-client-cert option" if a client certificate has
322 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
323 authenticators. This can be used for authorization after authentication
324 succeeds. (In the case of plaintext, it servers for both authentication
327 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
328 if any retry times were supplied.
330 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
331 connection, where there is no client IP address to check. In this
332 situation, the verify now always succeeds.
334 PH/25 Applied John Jetmore's -Mset patch.
336 PH/26 Added -bem to be like -Mset, but loading a message from a file.
338 PH/27 In a string expansion for a processed (not raw) header when multiple
339 headers of the same name were present, leading whitespace was being
340 removed from all of them, but trailing whitespace was being removed only
341 from the last one. Now trailing whitespace is removed from each header
342 before concatenation. Completely empty headers in a concatenation (as
345 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
346 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
348 PH/29 [Removed. This was a change that I later backed out, and forgot to
349 correct the ChangeLog entry (that I had efficiently created) before
350 committing the later change.]
352 PH/30 Exim was sometimes attempting to deliver messages that had suffered
353 address errors (4xx response to RCPT) over the same connection as other
354 messages routed to the same hosts. Such deliveries are always "forced",
355 so retry times are not inspected. This resulted in far too many retries
356 for the affected addresses. The effect occurred only when there were more
357 hosts than the hosts_max_try setting in the smtp transport when it had
358 the 4xx errors. Those hosts that it had tried were not added to the list
359 of hosts for which the message was waiting, so if all were tried, there
360 was no problem. Two fixes have been applied:
362 (i) If there are any address or message errors in an SMTP delivery, none
363 of the hosts (tried or untried) are now added to the list of hosts
364 for which the message is waiting, so the message should not be a
365 candidate for sending over the same connection that was used for a
366 successful delivery of some other message. This seems entirely
367 reasonable: after all the message is NOT "waiting for some host".
368 This is so "obvious" that I'm not sure why it wasn't done
369 previously. Hope I haven't missed anything, but it can't do any
370 harm, as the worst effect is to miss an optimization.
372 (ii) If, despite (i), such a delivery is accidentally attempted, the
373 routing retry time is respected, so at least it doesn't keep
374 hammering the server.
376 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
377 in ${readsocket because some servers need this prod.
379 PH/32 Added some extra debug output when updating a wait-xxx database.
381 PH/33 The hint "could be header name not terminated by colon", which has been
382 given for certain expansion errors for a long time, was not being given
383 for the ${if def:h_colon_omitted{... case.
385 PH/34 The spec says: "With one important exception, whenever a domain list is
386 being scanned, $domain contains the subject domain." There was at least
387 one case where this was not true.
389 PH/35 The error "getsockname() failed: connection reset by peer" was being
390 written to the panic log as well as the main log, but it isn't really
391 panic-worthy as it just means the connection died rather early on. I have
392 removed the panic log writing for the ECONNRESET error when getsockname()
395 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
396 runs only) independently of the message's sender address. This meant
397 that, if the 4xx error was in fact related to the sender, a different
398 message to the same recipient with a different sender could confuse
399 things. In particualar, this can happen when sending to a greylisting
400 server, but other circumstances could also provoke similar problems.
401 I have changed the default so that the retry time for these errors is now
402 based a combination of the sender and recipient addresses. This change
403 can be overridden by setting address_retry_include_sender=false in the
406 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
407 remote server are returned as part of bounce messages. This was not
408 happening for LMTP over a pipe (the lmtp transport), but now it is the
409 same for both kinds of LMTP.
411 PH/38 Despite being documented as not happening, Exim was rewriting addresses
412 in header lines that were in fact CNAMEs. This is no longer the case.
414 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
415 and queue runs started by the daemon processed all messages. This has
416 been fixed so that -R and -S can now usefully be given with -q<time>.
418 PH/40 Import PCRE release 6.7 (fixes some bugs).
420 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
422 PH/42 Give an error if -q is specified more than once.
424 PH/43 Renamed the variables $interface_address and $interface_port as
425 $received_ip_address and $received_port, to make it clear that these
426 values apply to message reception, and not to the outgoing interface when
427 a message is delivered. (The old names remain recognized, of course.)
429 PH/44 There was no timeout on the connect() call when using a Unix domain
430 socket in the ${readsocket expansion. There now is.
432 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
433 be meaningful with "accept".
435 SC/01 Eximstats V1.43
436 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
438 SC/02 Eximstats V1.44
439 Use a glob alias rather than an array ref in the generated
440 parser. This improves both readability and performance.
442 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
443 Collect SpamAssassin and rejection statistics.
444 Don't display local sender or destination tables unless
445 there is data to show.
446 Added average volumes into the top table text output.
448 SC/04 Eximstats V1.46
449 Collect data on the number of addresses (recipients)
450 as well as the number of messages.
452 SC/05 Eximstats V1.47
453 Added 'Message too big' to the list of mail rejection
454 reasons (thanks to Marco Gaiarin).
456 SC/06 Eximstats V1.48
457 Mainlog lines which have GMT offsets and are too short to
458 have a flag are now skipped.
460 SC/07 Eximstats V1.49 (Alain Williams)
461 Added the -emptyok flag.
463 SC/08 Eximstats V1.50
464 Fixes for obtaining the IP address from reject messages.
466 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
467 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
468 whitesspace changes from 4.64-PH/27
470 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
473 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
474 are found, allow negative numbers in numeric criteria)
476 JJ/06 exipick.20061117.2, added new $message_body_missing variable
478 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
479 to match changes made in 4.64-PH/43
481 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
483 PH/47 Put in an explicit test for a DNS lookup of an address record where the
484 "domain" is actually an IP address, and force a failure. This locks out
485 those revolvers/nameservers that support "A-for-A" lookups, in
486 contravention of the specifications.
488 PH/48 When a host name was looked up from an IP address, and the subsequent
489 forward lookup of the name timed out, the host name was left in
490 $sender_host_name, contrary to the specification.
492 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
493 restricted to single-key lookups, Exim was not diagnosing an error if
494 * or *@ was used with a query-style lookup.
496 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
498 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
499 long ago, but noone interested enough thought of it. Let's just say that
500 the "1.1" means that there are some new functions that weren't there at
501 some point in the past.
503 PH/51 Error processing for expansion failure of helo_data from an smtp
504 transport during callout processing was broken.
506 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
507 tested/used via the -bh/-bhc/-bs options.
509 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
510 bug, fixed in subsequent PCRE releases).
512 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
513 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
515 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
521 SC/01 Use a glob alias rather than an array ref in eximstats generated
522 parser. This improves both readability and performance.
524 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
525 Don't display local sender or destination tables in eximstats unless
526 there is data to show.
527 Added average volumes into the eximstats top table text output.
529 SC/03 Collect data on the number of addresses (recipients) as well
530 as the number of messages in eximstats.
532 TF/01 Correct an error in the documentation for the redirect router. Exim
533 does (usually) call initgroups() when daemonizing.
535 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
536 with consistent privilege compared to when running as a daemon.
538 TF/03 Note in the spec that $authenticated_id is not set for local
539 submissions from trusted users.
541 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
542 Thanks to Dean Brooks <dean@iglou.com> for the patch.
544 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
545 by adding some example configuration directives to the default
546 configuration file. A little bit of work is required to uncomment the
547 directives and define how usernames and passwords are checked, but
548 there is now a framework to start from.
550 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
551 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
552 without this. I don't know how relevant this is to other LDAP libraries.
554 PH/02 Add the verb name to the "unknown ACL verb" error.
556 PH/03 Magnus Holmgren's patch for filter_prepend_home.
558 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
560 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
561 directory not expanded when it should be if an expanded home directory
562 was set for the address (which is overridden by the transport).
564 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
567 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
568 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
569 because it is too late at that time, and has no effect.
571 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
572 security issue with \' (bugzilla #107). I could not use the
573 PQescapeStringConn() function, because it needs a PGconn value as one of
576 PH/08 When testing addresses using -bt, indicate those final addresses that
577 are duplicates that would not cause an additional delivery. At least one
578 person was confused, thinking that -bt output corresponded to deliveries.
579 (Suppressing duplicates isn't a good idea as you lose the information
580 about possibly different redirections that led to the duplicates.)
582 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
583 systems where poll() doesn't work, in particular OS X.
585 PH/10 Added more information to debugging output for retry time not reached.
587 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
588 operations in malware.c.
590 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
593 PH/13 If write_rejectlog was set false when logging was sent to syslog with
594 syslog_duplication set false, log lines that would normally be written
595 both the the main log and to the reject log were not written to syslog at
598 PH/14 In the default configuration, change the use of "message" in ACL warn
599 statements to "add_header".
601 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
602 not followed by a command (e.g. "seen endif").
604 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
605 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
608 PH/17 Added extra conditions to the default value of delay_warning_condition
612 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
613 { match{$h_precedence:}{(?i)bulk|list|junk} } \
614 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
617 The Auto-Submitted: and various List- headers are standardised, whereas I
618 don't think Precedence: ever was.
620 PH/18 Refactored debugging code in route_finduser() to show more information,
621 in particular, the error code if getpwnam() issues one.
623 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
624 This is apparently needed in addition to the PH/07 change above to avoid
625 any possible encoding problems.
627 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
628 but not after initializing Perl.
630 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
631 output them only if debugging. By default they are written stderr,
632 apparently, which is not desirable.
634 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
637 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
640 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
642 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
643 authenticated or an ident call has been made. Suppress the default
644 values for $authenticated_id and $authenticated_sender (but permit -oMai
645 and -oMas) when testing with -bh.
647 PH/24 Re-jigged the order of the tests in the default configuration so that the
648 tests for valid domains and recipients precede the DNS black list and CSA
649 tests, on the grounds that those ones are more expensive.
651 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
652 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
653 This bug exists in every version of Exim that I still have, right back to
656 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
657 However, an attempt to turn on case-sensitivity in a regex key by
658 including (?-i) didn't work because the subject string was already
659 lowercased, and the effects were non-intuitive. It turns out that a
660 one-line patch can be used to allow (?-i) to work as expected.
666 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
667 other effects) broke the use of negated acl sub-conditions.
669 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
672 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
673 "Deny" causes Exim to reject the incoming connection with a 554 error.
674 Unfortunately, if there is a major crisis, such as a disk failure,
675 tcp-wrappers gives "deny", whereas what one would like would be some
676 kind of temporary error. A kludge has been added to help with this.
677 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
678 554 error is used if errno is still zero or contains ENOENT (which occurs
679 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
682 PH/03 Add -lutil to the default FreeBSD LIBS setting.
684 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
685 errors. Otherwise a message that provokes a temporary error (when other
686 messages do not) can cause a whole host to time out.
688 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
689 addresses were routed directly to files or pipes from a redirect router.
690 File deliveries just didn't batch; pipe deliveries might have suffered
693 PH/06 A failure to get a lock for a hints database would erroneously always say
694 "Failed to get write lock", even when it was really a read lock.
696 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
697 of 0600. This has been changed to use the value of the lockfile_mode
698 option (which defaults to 0600).
700 PH/08 Applied small patch from the Sieve maintainer.
702 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
703 folder from quota calculations, a direct delivery into this folder messed
704 up the contents of the maildirsize file. This was because the regex was
705 used only to exclude .Trash (or whatever) when the size of the mailbox
706 was calculated. There was no check that a delivery was happening into an
707 excluded directory. This bug has been fixed by ignoring all quota
708 processing for deliveries into excluded directories.
710 PH/10 Added the maildirfolder_create_regex option to appendfile.
716 PH/01 The code for finding all the local interface addresses on a FreeBSD
717 system running IPv6 was broken. This may well have applied to all BSD
718 systems, as well as to others that have similar system calls. The broken
719 code found IPv4 interfaces correctly, but gave incorrect values for the
720 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
721 that it would not match correctly against @[] and not recognize the IPv6
724 PH/02 The ipliteral router was not recognizing addresses of the form user@
725 [ipv6:....] because it didn't know about the "ipv6:" prefix.
727 PH/03 Added disable_ipv6.
729 PH/04 Changed $reply_address to use the raw form of the headers instead of the
730 decoded form, because it is most often used to construct To: headers
731 lines in autoreplies, and the decoded form may well be syntactically
732 invalid. However, $reply_address has leading white space removed, and all
733 newlines turned into spaces so that the autoreply transport does not
736 PH/05 If group was specified without a user on a router, and no group or user
737 was specified on a transport, the group from the router was ignored.
739 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
740 for visible compile-time settings that can be used to change these
741 numbers, for those that want even more. Backwards compatibility with old
742 spool files has been maintained. However, going back to a previous Exim
743 release will lost any variables that are in spool files.
745 PH/07 Two small changes when running in the test harness: increase delay when
746 passing a TCP/IP connection to a new process, in case the original
747 process has to generate a bounce, and remove special handling of
748 127.0.0.2 (sic), which is no longer necessary.
750 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
751 be the same on different OS.
753 PH/09 Moved a debug statement in filter processing to avoid a race problem when
756 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
757 whether --show-vars was specified or not
759 JJ/02 exipick: Added support for new ACL variable spool format introduced
762 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
763 syntactically invalid From: or Reply-to: line, and a filter used this to
764 generate an autoreply, and therefore failed to obtain an address for the
765 autoreply, Exim could try to deliver to a non-existent relative file
766 name, causing unrelated and misleading errors. What now happens is that
767 it logs this as a hard delivery error, but does not attempt to create a
770 PH/11 The exinext utility has a -C option for testing purposes, but although
771 the given file was scanned by exinext itself; it wasn't being passed on
772 when Exim was called.
774 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
775 an end-of-file indication when reading a command response.
777 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
778 compiled. In many other places in Exim, IPv6 addresses are always
779 recognized, so I have changed this. It also means that IPv4 domain
780 literals of the form [IPV4:n.n.n.n] are now always recognized.
782 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
783 used if the router is not running as root, for example, when verifying at
784 ACL time, or when using -bh. The debugging output from this situation was
785 non-existent - all you got was a failure to exec. I have made two
788 (a) Failures to set uid/gid, the current directory, or a process leader
789 in a subprocess such as that created by queryprogram now generate
790 suitable debugging ouput when -d is set.
792 (b) The queryprogram router detects when it is not running as root,
793 outputs suitable debugging information if -d is set, and then runs
794 the subprocess without attempting to change uid/gid.
796 PH/15 Minor change to Makefile for building test_host (undocumented testing
799 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
800 additional section of a DNS packet that returns MX or SRV records.
801 Instead, it always explicitly searches for A/AAAA records. This avoids
802 major problems that occur when a DNS server includes only records of one
803 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
804 fixed another bug: if SRV records were looked up and the corresponding
805 address records were *not* found in the additional section, the port
806 values from the SRV records were lost.
808 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
809 using the correct key (the original address) when searching the retry
810 rules in order to find which one to use for generating the retry hint.
812 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
813 adding the default one. Similarly, if it contains a Reply-To: header, the
814 errors_reply_to option, if set, is not used.
816 PH/19 When calculating a retry time, Exim used to measure the "time since
817 failure" by looking at the "first failed" field in the retry record. Now
818 it does not use this if it is later than than the arrival time of the
819 message. Instead it uses the arrival time. This makes for better
820 behaviour in cases where some deliveries succeed, thus re-setting the
821 "first failed" field. An example is a quota failure for a huge message
822 when small messages continue to be delivered. Without this change, the
823 "time since failure" will always be short, possible causing more frequent
824 delivery attempts for the huge message than are intended.
825 [Note: This change was subsequently modified - see PH/04 for 4.62.]
827 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
828 $1, $2, $3) because the numerical variables can be reset during some
829 expansion items (e.g. "match"), thereby losing the authentication data.
831 PH/21 Make -bV show the size of off_t variables so that the test suite can
832 decide whether to run tests for quotas > 2G.
834 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
835 mailbox_size, and mailbox_filecount in the appendfile transport. If a
836 filecount value is greater than 2G or if a quota value is greater than 2G
837 on a system where the size of off_t is not greater than 4, a panic error
840 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
841 never match. The debug and -bh output now contains an explicit error
842 message indicating a malformed IPv4 address or mask.
844 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
845 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
848 PH/25 Do not write to syslog when running in the test harness. The only
849 occasion when this arises is a failure to open the main or panic logs
850 (for which there is an explicit test).
852 PH/26 Added the /no_tell option to "control=freeze".
854 PH/27 If a host name lookup failed very early in a connection, for example, if
855 the IP address matched host_lookup and the reverse lookup yielded a name
856 that did not have a forward lookup, an error message of the form "no IP
857 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
858 could be logged. Now it outputs the IP address instead of "NULL".
860 PH/28 An enabling patch from MH: add new function child_open_exim2() which
861 allows the sender and the authenticated sender to be set when
862 submitting a message from within Exim. Since child_open_exim() is
863 documented for local_scan(), the new function should be too.
865 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
866 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
867 results in an empty string is now treated as unset.
869 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
871 PH/31 Added sender_verify_fail logging option.
873 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
874 needed by RFC 822 but not by RFC 2822 was commented out. I have now
875 tidied the source and removed it altogether.
877 PH/33 When a queue run was abandoned because the load average was too high, a
878 log line was always written; now it is written only if the queue_run log
879 selector is set. In addition, the log line for abandonment now contains
880 information about the queue run such as the pid. This is always present
881 in "start" and "stop" lines but was omitted from the "abandon" line.
883 PH/34 Omit spaces between a header name and the colon in the error message that
884 is given when verify = headers_syntax fails (if there are lots of them,
885 the message gets confusing).
887 PH/35 Change the default for dns_check_names_pattern to allow slashes within
888 names, as there are now some PTR records that contain slashes. This check
889 is only to protect against broken name servers that fall over on strange
890 characters, so the fact that it applies to all lookups doesn't matter.
892 PH/36 Now that the new test suite is complete, we can remove some of the
893 special code in Exim that was needed for the old test suite. For example,
894 sorting DNS records because real resolvers return them in an arbitrary
895 order. The new test suite's fake resolver always returns records in the
898 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
899 bounces) except when queue_only is set, to avoid logging races between
900 the different processes.
902 PH/38 Panic-die if .include specifies a non-absolute path.
904 PH/39 A tweak to the "H" retry rule from its user.
906 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
907 a label. They prevented compilation on older perls.
909 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
910 a warning to be raised on newish perls.
912 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
913 on queue. Changes to match documented behaviour of showing count of
914 messages matching specified criteria.
916 PH/40 Changed the default ident timeout from 30s to 5s.
918 PH/41 Added support for the use of login_cap features, on those BSD systems
919 that have them, for controlling the resources used by pipe deliveries.
921 PH/42 The content-scanning code uses fopen() to create files in which to put
922 message data. Previously it was not paying any attention to the mode of
923 the files. Exim runs with umask(0) because the rest of the code creates
924 files with open(), and sets the required mode explicitly. Thus, these
925 files were ending up world-writeable. This was not a big issue, because,
926 being within the spool directory, they were not world-accessible. I have
927 created a function called modefopen, which takes an additional mode
928 argument. It sets umask(777), creates the file, chmods it to the required
929 mode, then resets the umask. All the relevant calls to fopen() in the
930 content scanning code have been changed to use this function.
932 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
933 to 24 hours. This avoids potential overflow problems when processing G
934 and H retry rules. I suspect nobody ever tinkers with this value.
936 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
938 PH/45 When the plaintext authenticator is running as a client, the server's
939 challenges are checked to ensure they are valid base64 strings. By
940 default, the authentication attempt is cancelled if an invalid string is
941 received. Setting client_ignore_invalid_base64 true ignores these errors.
942 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
943 they are received. Thus, the responses can be made to depend on the
944 challenges. If an invalid string is ignored, an empty string is placed in
947 PH/46 Messages that are created by the autoreply transport now contains a
948 References: header, in accordance with RFCs 2822 and 3834.
950 PH/47 Added authenticated_sender_force to the smtp transport.
952 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
954 PH/49 Installed latest patch from the Sieve maintainer.
956 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
957 was also set, the appendfile transport was unnecessarily scanning a
958 directory of message files (e.g. for maildir delivery) to find the count
959 of files (along with the size), even though it did not need this
960 information. It now does the scan only if it needs to find either the
961 size of the count of files.
963 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
965 PH/52 Two bugs concerned with error handling when the smtp transport is
968 (i) Exim was not creating retry information for temporary errors given
969 for individual recipients after the DATA command when the smtp transport
970 was used in LMTP mode. This meant that they could be retried too
971 frequently, and not timed out correctly.
973 (ii) Exim was setting the flag that allows error details to be returned
974 for LMTP errors on RCPT commands, but not for LMTP errors for individual
975 recipients that were returned after the DATA command.
977 PH/53 This is related to PH/52, but is more general: for any failing address,
978 when detailed error information was permitted to be returned to the
979 sender, but the error was temporary, then after the final timeout, only
980 "retry timeout exceeded" was returned. Now it returns the full error as
981 well as "retry timeout exceeded".
983 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
984 do this, and (what is worse) MTAs that accept it.
986 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
987 will now be deprecated.
989 PH/56 New os.c-cygwin from the Cygwin maintainer.
991 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
992 formats (previously only available in exim formats via -bpr, -bpru,
993 and -bpra. Now also available in native and exiqgrep formats)
995 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
996 with very large, slow to parse queues
998 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1000 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1002 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1003 responses to authentication challenges, though it was showing the
1004 challenges; (ii) I've removed the CR characters from the debug output for
1007 PH/58 Allow for the insertion of a newline as well as a space when a string
1008 is turned into more than one encoded-word during RFC 2047 encoding. The
1009 Sieve code now uses this.
1011 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1012 data_4xx, lost_connection, tls_required.
1014 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1015 message was being sent as an SMTP response.
1017 PH/61 Add -l and -k options to exicyclog.
1019 PH/62 When verifying, if an address was redirected to one new address, so that
1020 verification continued, and the new address failed or deferred after
1021 having set something in $address_data, the value of $address_data was not
1022 passed back to the ACL. This was different to the case when no
1023 redirection occurred. The value is now passed back in both cases.
1025 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1026 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1027 use login_cap.h, so on its own it isn't the distinguishing feature. The
1028 new name refers directly to the setclassresources() function.
1030 PH/65 Added configuration files for NetBSD3.
1032 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1034 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1035 is preferred over IPv4.
1037 PH/68 The bounce_return_message and bounce_return_body options were not being
1038 honoured for bounces generated during the reception of non-SMTP messages.
1039 In particular, this applied to messages rejected by the ACL. This bug has
1040 been fixed. However, if bounce_return_message is true and bounce_return_
1041 body is false, the headers that are returned for a non-SMTP message
1042 include only those that have been read before the error was detected.
1043 (In the case of an ACL rejection, they have all been read.)
1045 PH/69 The HTML version of the specification is now built in a directory called
1046 spec_html instead of spec.html, because the latter looks like a path with
1047 a MIME-type, and this confuses some software.
1049 PH/70 Catch two compiler warnings in sieve.c.
1051 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1052 function verify_get_ident() calls ip_connect() to connect a socket, but
1053 if the "connect()" function timed out, ip_connect() used to close the
1054 socket. However, verify_get_ident() also closes the socket later, and in
1055 between Exim writes to the log, which may get opened at this point. When
1056 the socket was closed in ip_connect(), the log could get the same file
1057 descriptor number as the socket. This naturally causes chaos. The fix is
1058 not to close the socket in ip_connect(); the socket should be closed by
1059 the function that creates it. There was only one place in the code where
1060 this was missing, in the iplookup router, which I don't think anybody now
1061 uses, but I've fixed it anyway.
1063 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1064 well as to direct DNS lookups. Otherwise the handling of names in host
1065 lists is inconsistent and therefore confusing.
1071 PH/01 Two changes to the default runtime configuration:
1073 (1) Move the checks for relay_from_hosts and authenticated clients from
1074 after to before the (commented out) DNS black list checks.
1076 (2) Add control=submission to the relay_from_hosts and authenticated
1077 clients checks, on the grounds that messages accepted by these
1078 statements are most likely to be submissions.
1080 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1082 (1) Generate an error if the third argument for the ${prvs expansion is
1085 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1088 (3) Reset the variables that are obtained from the first argument of
1089 ${prvscheck and used in the second argument before leaving the code,
1090 because their memory is reclaimed, so using them afterwards may do
1093 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1094 one (it's much easier than Tom thought :-).
1096 (5) Because of (4), we can now allow for the use of $prvscheck_result
1097 inside the third argument.
1099 PH/03 For some reason, the default setting of PATH when running a command from
1100 a pipe transport was just "/usr/bin". I have changed it to
1103 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1104 anything to be listed in the output from -bV.
1106 PH/05 When a filter generated an autoreply, the entire To: header line was
1107 quoted in the delivery log line, like this:
1109 => >A.N.Other <ano@some.domain> <original@ddress> ...
1111 This has been changed so that it extracts the operative address. There
1112 may be more than one such address. If so, they are comma-separated, like
1115 => >ano@some.domain,ona@other.domain <original@ddress> ...
1117 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1118 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1119 not being looked up in the rDNS to get a host name, Exim was showing the
1120 IP address twice in Received: lines, even though the IP addresses were
1121 identical. For example:
1123 Received: from [1.2.3.4] (helo=[1.2.3.4])
1125 However, if the real host name was known, it was omitting the HELO data
1126 if it matched the actual IP address. This has been tidied up so that it
1127 doesn't show the same IP address twice.
1129 PH/07 When both +timestamp and +memory debugging was on, the value given by
1130 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1131 called by the debug printing, thereby overwriting the timestamp buffer.
1132 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1134 PH/08 When the original message was included in an autoreply transport, it
1135 always said "this is a copy of the message, including all the headers",
1136 even if body_only or headers_only was set. It now gives an appropriate
1139 PH/09 Applied a patch from the Sieve maintainer which:
1141 o fixes some comments
1142 o adds the (disabled) notify extension core
1143 o adds some debug output for the result of if/elsif tests
1144 o points to the current vacation draft in the documentation
1145 and documents the missing references header update
1149 o fixes a bug in processing the envelope test (when testing
1150 multiple envelope elements, the last element determinted the
1153 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1154 Electronic Mail") by including:
1156 Auto-submitted: auto-generated
1158 in the messages that it generates (bounce messages and others, such as
1159 warnings). In the case of bounce messages for non-SMTP mesages, there was
1160 also a typo: it was using "Auto_submitted" (underscore instead of
1161 hyphen). Since every message generated by Exim is necessarily in response
1162 to another message, thes have all been changed to:
1164 Auto-Submitted: auto-replied
1166 in accordance with these statements in the RFC:
1168 The auto-replied keyword:
1170 - SHOULD be used on messages sent in direct response to another
1171 message by an automatic process,
1173 - MUST NOT be used on manually-generated messages,
1175 - MAY be used on Delivery Status Notifications (DSNs) and Message
1176 Disposition Notifications (MDNs),
1178 - MUST NOT be used on messages generated by automatic or periodic
1179 processes, except for messages which are automatic responses to
1182 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1183 to the default Received: header definition.
1185 PH/12 Added log selector acl_warn_skipped (default on).
1187 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1188 variables because (a) they are in the wrong storage pool and (b) even if
1189 they were copied, it wouldn't work properly because of the caching.
1191 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1192 checking when decoding. Apparently there are clients that generate
1193 overlong encoded strings. Why am I not surprised?
1195 PH/15 If the first argument of "${if match_address" was not empty, but did not
1196 contain an "@" character, Exim crashed. Now it writes a panic log message
1197 and treats the condition as false.
1199 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1201 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1202 extension "envelope-auth". The code is finished and in agreement with
1203 other implementations, but there is no documentation so far and in fact,
1204 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1205 not changing the active code.
1207 Print executed "if" and "elsif" statements when debugging is used. This
1208 helps a great deal to understand what a filter does.
1210 Document more things not specified clearly in RFC3028. I had all this
1211 sorted out, when out of a sudden new issues came to my mind. Oops."
1213 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1216 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1217 canonical form (as documented). However, after a host name lookup from
1218 the IP address, check_host() was doing a simple string comparison with
1219 addresses acquired from the DNS when checking that the found name did
1220 have the original IP as one of its addresses. Since any found IPv6
1221 addresses are likely to be in abbreviated form, the comparison could
1222 fail. Luckily, there already exists a function for doing the comparison
1223 by converting both addresses to binary, so now that is used instead of
1224 the text comparison.
1226 PH/20 There was another similar case to PH/19, when a complete host name was
1227 given in a host list; looking up its IP address could give an abbreviated
1228 form, whereas the current host's name might or might not be abbreviated.
1229 The same fix has been applied.
1235 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1236 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1239 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1240 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1242 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1244 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1245 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1246 Also, for "Auto-Submitted", treat anything other than "no" as
1247 non-personal, in accordance with RFC 3834. (Previously it treated
1248 anything starting "auto-" as non-personal.)
1250 TF/01 The control=submission/name=... option had a problem with syntax
1251 errors if the name included a slash character. The /name= option
1252 now slurps the rest of the string, so it can include any characters
1253 but it must come last in the list of options (after /sender_retain
1256 PH/05 Some modifications to the interface to the fake nameserver for the new
1264 TK/01 Added the "success_on_redirect" address verification option. See
1265 NewStuff for rationale and an example.
1267 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1269 PH/02 Patch to exigrep to allow it to work on syslog lines.
1271 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1272 fread() to skip over the body file's header line, because in Cygwin the
1273 header line is locked and is inaccessible.
1275 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1276 co-exist for some time) to make it clear that it is the Exim ID that is
1277 referenced, not the Message-ID: header line.
1279 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1280 string_format() function, because snprintf() does not exist on all
1283 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1284 ${stat: expansion item.
1286 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1287 protocol synchronization error", to keep the pedants happy.
1289 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1290 well as for IRIX systems, when gcc is being used. See the host.c source
1293 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1295 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1298 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1299 transport and to the smtp transport in LMTP mode.
1301 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1303 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1304 run a filter in a subprocess. This could lead to confusion in subsequent
1305 lookups in the parent process. There should also be a search_tidyup() at
1306 the end of the subprocess.
1308 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1309 only if the host matched helo_try_verify_hosts, which caused the
1310 verification to occur when the EHLO/HELO command was issued. The ACL just
1311 tested the remembered result. Now, if a previous verification attempt has
1312 not happened, "verify = helo" does it there and then.
1314 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1316 TK/03 Fix log output including CR from clamd.
1318 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1319 exist provoked a memory error which could cause a segfault.
1321 PH/15 Installed PCRE 6.2
1323 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1325 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1326 of the problem. Specifically, suggested +O2 rather than +O1 for the
1329 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1331 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1332 record, the port was not correctly logged when the outgoing_port log
1333 selector was set (it logged the transort's default port).
1335 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1336 fallback_hosts, and "hosts" in the smtp transport.
1338 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1339 host errors such as "Connection refused".
1341 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1342 authentication with radiusclient 0.4.9:
1344 - Error returned from rc_read_config was caught wrongly
1345 - Username/password not passed on to radius server due to wrong length.
1347 The presumption is that some radiusclient API changes for 4.51/PH/17
1348 were not taken care of correctly. The code is still untested by me (my
1349 Linux distribution still has 0.3.2 of radiusclient), but it was
1350 contributed by a Radius user.
1352 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1353 expanding the "port" option of the smtp transport.
1355 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1356 while reading a MIME header. Thanks to Tom Hughes for a patch.
1358 PH/24 Include config.h inside local_scan.h so that configuration settings are
1361 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1362 that in an ACL for RCPT (for example), you can examine exactly what was
1365 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1366 commands, but it was not correctly comparing the address with the actual
1367 client host address. Thus, it would show the EHLO address in Received:
1368 header lines when this was not necessary.
1370 PH/27 Added the % operator to ${eval:}.
1372 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1373 it should be ignoring failures (because with -C, for example, it has lost
1374 privilege). It wasn't ignoring creation failures other than "already
1377 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1380 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1381 because an input line was too long, either on its own, or by virtue of
1382 too many continuations, the temporary file was not being removed, and the
1383 return code was incorrect.
1385 PH/31 Missing "BOOL" in function definition in filtertest.c.
1387 PH/32 Applied Sieve patches from the maintainer.
1389 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1391 PH/33 Added "verify = not_blind".
1393 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1394 Local/Makefile (with some defaults set). These are used in built scripts
1395 such as exicyclog, but they have never been used in the exim_install
1396 script (though there are many overriding facilities there). I have
1397 arranged that the exim_install script now takes note of these two
1400 PH/35 Installed configuration files for Dragonfly.
1402 PH/36 When a locally submitted message by a trusted user did not contain a
1403 From: header, and the sender address was obtained from -f or from an SMTP
1404 MAIL command, and the trusted user did not use -F to supply a sender
1405 name, $originator_name was incorrectly used when constructing a From:
1406 header. Furthermore, $originator_name was used for submission mode
1407 messages from external hosts without From: headers in a similar way,
1408 which is clearly wrong.
1410 PH/37 Added control=suppress_local_fixups.
1412 PH/38 When log_selector = +received_sender was set, and the addition of the
1413 sender made the log line's construction buffer exactly full, or one byte
1414 less than full, an overflow happened when the terminating "\n" was
1417 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1418 when the result of a list match is failure because a DNS lookup failed.
1420 PH/40 RM_COMMAND is now used in the building process.
1422 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1423 the "build-* directories that it finds.
1425 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1426 address in a domain literal was a prefix of an interface address.
1428 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1429 when verifying a sender address, unless rewrite_headers is false.
1431 PH/44 Wrote a long comment about why errors_to addresses are verified as
1432 recipients, not senders.
1434 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1435 the ratelimit ACL was added.
1437 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1439 PH/46 Added extra information about PostgreSQL errors to the error string.
1441 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1442 avoiding the need to install special zones in a real server. This is
1443 backwards compatible; if it can't find the fake resolver, it drops back.
1444 Thus, both old and new test suites can be run.
1446 TF/02 Added util/ratelimit.pl
1448 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1449 clock is set back in time.
1451 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1452 Candler <B.Candler@pobox.com>.
1454 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1455 OK for addresses that are the result of redirections.
1457 PH/48 A number of further additions for the benefit of the new test suite,
1458 including a fake gethostbyname() that interfaces to the fake DNS resolver
1461 TF/06 The fix for widen_domains has also been applied to qualify_single and
1462 search_parents which are the other dnslookup options that can cause
1465 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1468 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1470 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1471 service" if it encountered an MX record with an empty target hostname.
1472 The message is now "an MX or SRV record indicated no SMTP service".
1474 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1475 if the DNS of the sending site is misconfigured. This is quite a
1476 common situation. This change restores the behaviour of treating a
1477 helo verification defer as a failure.
1479 PH/51 If self=fail was set on a router, the bounce message did not include the
1480 actual error message.
1486 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1488 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1489 command itself ended in error, the underlying message about the transport
1490 filter timeout was being overwritten with the pipe command error. Now the
1491 underlying error message should be appended to the second error message.
1493 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1495 PH/02 Reduce the amount of output that "make" produces by default. Full output
1496 can still be requested.
1498 PH/03 The warning log line about a condition test deferring for a "warn" verb
1499 was being output only once per connection, rather than after each
1500 occurrence (because it was using the same function as for successful
1501 "warn" verbs). This seems wrong, so I have changed it.
1503 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1504 it should not have, which might have caused a crash in the right
1505 circumstances, but probably never did.
1507 PH/04 Installed a modified version of Tony Finch's patch to make submission
1508 mode fix the return path as well as the Sender: header line, and to
1509 add a /name= option so that you can make the user's friendly name appear
1512 TF/03 Added the control = fakedefer ACL modifier.
1514 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1515 Mark Lowes for thorough testing.
1517 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1519 TK/03 Merged latest SRS patch from Miles Wilton.
1521 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1522 with the definition in sysexits.h (which is #included earlier).
1523 Fortunately, Exim does not actually use EX_OK. The code used to try to
1524 preserve the sysexits.h value, by assumimg that macro definitions were
1525 scanned for macro replacements. I have been disabused of this notion,
1526 so now the code just undefines EX_OK before #including unistd.h.
1528 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1529 in the smtp transport. When a block could not be written in a single
1530 write() function, the timeout was being re-applied to each part-write.
1531 This seems wrong - if the receiver was accepting one byte at a time it
1532 would take for ever. The timeout is now adjusted when this happens. It
1533 doesn't have to be particularly precise.
1535 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1536 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1538 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1539 without a domain if the check to <postmaster@domain> fails.
1541 SC/01 Eximstats: added -xls and the ability to specify output files
1542 (patch written by Frank Heydlauf).
1544 SC/02 Eximstats: use FileHandles for outputing results.
1546 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1548 SC/04 Eximstats: fixed display of large numbers with -nvr option
1550 SC/05 Eximstats: fixed merging of reports with empty tables.
1552 SC/06 Eximstats: added the -include_original_destination flag
1554 SC/07 Eximstats: removed tabs and trailing whitespace.
1556 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1558 TK/06 MBOX spool code: Add real "From " MBOX separator line
1559 so the .eml file is really in mbox format (even though
1560 most programs do not really care). Patch from Alex Miller.
1562 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1563 The latter is generated from $received_to and is only set if the
1564 message has one envelope recipient. SA can use these headers,
1565 obviously out-of-the-box. Patch from Alex Miller.
1567 PH/08 The ${def test on a variable was returning false if the variable's
1568 value was "0", contrary to what the specification has always said!
1569 The result should be true unless the variable is empty.
1571 PH/09 The syntax error of a character other than { following "${if
1572 def:variable_name" (after optional whitespace) was not being diagnosed.
1573 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1574 accidental colon was present, for example, could give incorrect results.
1576 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1577 result is used (not including appendfile, where other changes are about
1580 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1581 This involved changing a lot of size variables from int to off_t. It
1582 should work with maildirs and everything.
1584 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1585 spamd dying while we are connected to it.
1587 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1590 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1591 function for BDB 4.3.
1593 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1595 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1596 the sources for PCRE have been re-arranged and more files are now
1599 PH/15 The code I had for printing potentially long long variables in PH/11
1600 above was not the best (it lost precision). The length of off_t variables
1601 is now inspected at build time, and an appropriate printing format (%ld
1602 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1603 to be "long long int" or "long int". This is needed for the internal
1604 formatting function string_vformat().
1606 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1607 the configuration file to be ":syslog", then the script "guesses" where
1608 the logs files are, rather than using the compiled in default. In our
1609 case the guess is not the same as the compiled default, so the script
1610 suddenly stopped working when I started to use syslog. The patch checks
1611 to see if log_file_path is "". If so, it attempts to read it from exim
1612 with no configuration file to get the compiled in version, before it
1613 falls back to the previous guessing code."
1615 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1616 implementing BATV in an Exim configuration. See NewStuff for the gory
1619 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1620 Makefile that are specific to HP-UX.
1622 PH/18 If the "use_postmaster" option was set for a recipient callout together
1623 with the "random" option, the postmaster address was used as the MAIL
1624 FROM address for the random test, but not for the subsequent recipient
1625 test. It is now used for both.
1627 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1628 patch removes a few documentation additions to RFC 3028, because the
1629 latest draft now contains them. It adds the new en;ascii-case comparator
1630 and a new error check for 8bit text in MIME parts. Comparator and
1631 require names are now matched exactly. I enabled the subaddress
1632 extension, but it is not well tested yet (read: it works for me)."
1634 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1635 rework some of the code of TK/09 above to avoid the hardwired use of
1636 "%lld" and "long long". Replaced the call to snprintf() with a call to
1639 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1640 records point to non-existent hosts", "retry timeout exceeded", and
1641 "retry time not reached for any host after a long failure period".
1643 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1644 experimental DomainKeys support:
1646 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1647 (2) On an error such as an illegally used "control", the wrong name for
1648 the control was given.
1650 These problems did NOT occur unless DomainKeys support was compiled.
1652 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1654 PH/24 Added ${if match_ip condition.
1656 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1657 databases so that it will be absolutely obvious if a crash occurs in the
1658 DB library. This is a regular occurrence (often caused by mis-matched
1661 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1662 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1663 picked up on a user's system that detects such things. There doesn't seem
1664 to be a gcc warning option for this - only an attribute that has to be
1665 put on the function's prototype. It seems that in Fedora Core 4 they have
1666 set this on a number of new functions. No doubt there will be more in due
1669 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1670 specify a transport. However, if an address that was verified by such a
1671 router was the subject of a callout, Exim crashed because it tried to
1672 read the rcpt_include_affixes from the non-existent transport. Now it
1673 just assumes that the setting of that option is false. This bug was
1674 introduced by 4.51/PH/31.
1676 PH/28 Changed -d+all to exclude +memory, because that information is very
1677 rarely of interest, but it makes the output a lot bigger. People tend to
1678 do -d+all out of habit.
1680 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1681 code in os-type was giving problems when libc.so lives in lib64, like on
1684 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1685 aren't the modern standard, and it seems that some systems' include files
1686 don't always have them. Exim was already checking for some of the newer
1687 ones like T_AAAA, and defining it itself. I've added checks for all the
1688 record types that Exim uses.
1690 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1691 not automatically generating a new one, as it is supposed to. This
1692 prevented TLS from working. If the file did exist, but contained invalid
1693 data, a new version was generated, as expected. It was only the case of a
1694 non-existent file that was broken.
1696 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1697 with a change in libdomainkeys > 0.64.
1699 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1700 from DNS. If the selector record carries the flag, it now has
1701 precedence over the domain-wide flag.
1703 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1705 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1706 the use of a port name that isn't defined in /etc/services) occurred, the
1707 message was deferred as in a normal delivery, and thus remained on the
1708 spool, instead of being failed because of the mua_wrapper setting. This
1709 is now fixed, and I tidied up some of the mua_wrapper messages at the
1712 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1713 the messages in a hash of arrays rather than using individual hashes.
1714 This is a bit cleaner and results in dramatic memory savings, albeit
1715 at a slight CPU cost.
1717 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1718 as requested by Marc Sherman.
1720 SC/10 Eximstats: added histograms for user specified patterns as requested
1723 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1725 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1726 fopen() in the content-scanning modules that did not already have it.
1732 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1733 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1735 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1737 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1739 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1740 to test Sieve filters that use "vacation".
1742 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1743 that changes the way the GnuTLS parameters are stored in the cache file.
1744 The new format can be generated externally. For backward compatibility,
1745 if the data in the cache doesn't make sense, Exim assumes it has read an
1746 old-format file, and it generates new data and writes a new file. This
1747 means that you can't go back to an older release without removing the
1750 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1751 work if there are any delivery delays because "one_time" forces the
1752 parent to be marked "delivered", so its unseen clone is never tried
1753 again. For this reason, Exim now forbids the simultaneous setting of
1756 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1757 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1758 entry if you want to know the details. The fix, however, neglected to
1759 consider the case when local delivery batching is involved. The test for
1760 "previously delivered" was not happening when checking to see if an
1761 address could be batched with a previous (undelivered) one; under
1762 certain circumstances this could lead to multiple deliveries to the same
1765 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1766 in its include files, and this causes problems building Exim.
1768 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1769 header_syntax) but Exim was just ignoring anything given after a slash.
1770 In particular, this caused confusion with an attempt to use "verify =
1771 reverse_host_lookup/defer_ok". An error is now given when options are
1772 supplied for verify items that do not have them. (Maybe reverse_host_
1773 lookup should have a defer_ok option, but that's a different point.)
1775 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1776 defined by RFC 821) to 2048, because there were problems with some AUTH
1777 commands, and RFC 1869 says the size should be increased for extended
1778 SMTP commands that take arguments.
1780 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1783 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1784 "unknown" error; now it says that the functionality isn't in the binary.
1786 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1787 an address' error message when a string expansion fails (syntax or
1788 whatever). Otherwise the password may appear in the log. Following change
1789 PH/42 below, there is no longer a chance of it appearing in a bounce
1792 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1794 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1795 of its addresses was ever tried. (Bugzilla bug #2.)
1797 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1798 the result incorrectly in the debug output. (It correctly added a newline
1799 to what was transported.)
1801 TF/01 Added $received_time.
1803 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1804 commented out examples of how to interface to a virus scanner and to
1805 SpamAssassin. Also added commented examples of av_scanner and
1806 spamd_address settings.
1808 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1809 and controls are allowed in which ACLs. There were a couple of minor
1810 errors. Some of the entries in the conditions table (which is a table of
1811 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1812 negation of where the condition IS allowed.
1814 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1816 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1817 header file does not have a version number, so I've had to invent a new
1818 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1819 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1820 radiusclient), but it was contributed by a Radius user.
1822 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1823 files or directories, for OpenSSL.
1825 PH/19 When an Exim process that is running as root has to create an Exim log
1826 file, it does so in a subprocess that runs as exim:exim so as to get the
1827 ownership right at creation (otherwise, other Exim processes might see
1828 the file with the wrong ownership). There was no test for failure of this
1829 fork() call, which would lead to the process getting stuck as it waited
1830 for a non-existent subprocess. Forks do occasionally fail when resources
1831 run out. I reviewed all the other calls to fork(); they all seem to check
1834 PH/20 When checking for unexpected SMTP input at connect time (before writing
1835 the banner), Exim was not dealing correctly with a non-positive return
1836 from the read() function. If the client had disconnected by this time,
1837 the result was a log entry for a synchronization error with an empty
1838 string after "input=" when read() returned zero. If read() returned -1
1839 (an event I could not check), uninitialized data bytes were printed.
1840 There were reports of junk text (parts of files, etc) appearing after
1843 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1845 PH/22 Added support for macro redefinition, and (re)definition in between
1846 driver and ACL definitions.
1848 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1849 forgetting to use the resulting value; it was using the unexpanded value.
1851 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1852 hadn't been configured. The fix is from Juergen Kreileder, who
1853 understands it better than I do:
1855 "Here's what I see happening with three configured cyrus_sasl
1856 authenticators configured (plain, login, cram-md5):
1858 On startup auth_cyrus_sasl_init() gets called for each of these.
1859 This means three calls to sasl_listmech() without a specified mech_list.
1860 => SASL tests which mechs of all available mechs actually work
1861 => three warnings about OTP not working
1862 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1864 With the patch, sasl_listmech() also gets called three times. But now
1865 SASL's mech_list option is set to the server_mech specified in the the
1866 authenticator. Or in other words, the answer from sasl_listmech()
1867 gets limited to just the mech you're testing for (which is different
1869 => the return list contains just 'plain' or 'login', 'cram-md5' or
1870 nothing depending on the value of ob->server_mech.
1872 I've just tested the patch: Authentication still works fine,
1873 unavailable mechs specified in the exim configuration are still
1874 caught, and the auth.log warnings about OTP are gone."
1876 PH/25 When debugging is enabled, the contents of the command line are added
1877 to the debugging output, even when log_selector=+arguments is not
1880 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1881 answer is "GNU", and only if the return is "GNU/something" is the answer
1884 PH/27 $acl_verify_message is now set immediately after the failure of a
1885 verification in an ACL, and so is available in subsequent modifiers. In
1886 particular, the message can be preserved by coding like this:
1888 warn !verify = sender
1889 set acl_m0 = $acl_verify_message
1891 Previously, $acl_verify_message was set only while expanding "message"
1892 and "log_message" when a very denied access.
1894 PH/28 Modified OS/os.c-Linux with
1896 -#ifndef OS_LOAD_AVERAGE
1897 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1899 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1900 nomenclature these days.)
1902 PH/29 Installed patch from the Sieve maintainer that adds the options
1903 sieve_useraddress and sieve_subaddress to the redirect router.
1905 PH/30 In these circumstances:
1906 . Two addresses routed to the same list of hosts;
1907 . First host does not offer TLS;
1908 . First host accepts first address;
1909 . First host gives temporary error to second address;
1910 . Second host offers TLS and a TLS session is established;
1911 . Second host accepts second address.
1912 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1913 and peerdn, if requested) that were in fact used only for the second
1916 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1917 attention to any local part prefix or suffix that was matched by the
1918 router that accepted the address. It now behaves in the same way as it
1919 does for delivery: the affixes are removed from the local part unless
1920 rcpt_include_affixes is set on the transport.
1922 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1923 timeout during the DATA phase of an incoming message.
1925 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1926 applied a patch sanctioned by the Sieve maintainer.
1928 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1929 the uid or gid is negative. A case of a negative gid caused this to be
1930 noticed. The fix allows for either to be negative.
1932 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1933 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1936 PH/36 The change PH/12 above was broken. Fixed it.
1938 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1939 the grounds that routing the same address twice would always produce the
1940 same answer. This might have been true once, but it is certainly no
1941 longer true now. Routing a child address may depend on the previous
1942 routing that produced that child. Some complicated redirection strategies
1943 went wrong when messages had multiple recipients, and made Exim's
1944 behaviour dependent on the order in which the addresses were given.
1946 I have moved the duplicate checking until after the routing is complete.
1947 Exim scans the addresses that are assigned to local and remote
1948 transports, and removes any duplicates. This means that more work will be
1949 done, as duplicates will always all be routed, but duplicates are
1950 presumably rare, so I don't expect this is of any significance.
1952 For deliveries to pipes, files, and autoreplies, the duplicate checking
1953 still happens during the routing process, since they are not going to be
1956 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1957 It corrects a timeout issue with spamd. This is Ian's comment: "The
1958 background is that sometimes spamd either never reads data from a
1959 connection it has accepted, or it never writes response data. The exiscan
1960 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1961 blindly assumes that writes won't block so it may never time out."
1963 PH/39 Allow G after quota size as well as K and M.
1965 PH/40 The value set for $authenticated_id in an authenticator may not contain
1966 binary zeroes or newlines because the value is written to log lines and
1967 to spool files. There was no check on this. Now the value is run through
1968 the string_printing() function so that such characters are converted to
1969 printable escape sequences.
1971 PH/41 $message_linecount is a new variable that contains the total number of
1972 lines in the message. Compare $body_linecount, which is the count for the
1975 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1976 bounce and delay warning messages, except in certain special cases, which
1979 (a) An SMTP error message from a remote host;
1980 (b) A message specified in a :fail: redirection;
1981 (c) A message specified in a "fail" command in a system filter;
1982 (d) A message specified in a FAIL return from the queryprogram router;
1983 (e) A message specified by the cannot_route_message router option.
1985 In these cases only, Exim does include the error details in bounce and
1986 warning messages. There are also a few cases where bland messages such
1987 as "unrouteable address" or "local delivery error" are given.
1989 PH/43 $value is now also set for the "else" part of a ${run expansion.
1991 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1992 being worked on, but at least Exim now implements the latest version to
1995 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1996 process to complete was treated as a delivery failure, a timeout while
1997 writing the message to the pipe was logged, but erroneously treated as a
1998 successful delivery. Such timeouts include transport filter timeouts. For
1999 consistency with the overall process timeout, these timeouts are now
2000 treated as errors, giving rise to delivery failures by default. However,
2001 there is now a new Boolean option for the pipe transport called
2002 timeout_defer, which, if set TRUE, converts the failures into defers for
2003 both kinds of timeout. A transport filter timeout is now identified in
2006 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2007 systems where "make" and "gmake" are different, calling "gmake" at top
2008 level broke things. I've arranged for the value of $(MAKE) to be passed
2009 from the Makefile to this script so that it can call the same version of
2013 A note about Exim versions 4.44 and 4.50
2014 ----------------------------------------
2016 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2017 changes of various kinds. As a consequence, a big documentation update was
2018 needed. This delayed the release for rather longer than seemed good, especially
2019 in the light of a couple of (minor) security issues. Therefore, the changes
2020 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2021 release. So 4.44 and 4.50 are in effect two different branches that both start
2024 I have left the 4.50 change log unchanged; it contains all the changes since
2025 4.43. The change log for 4.44 is below; many of its items are identical to
2026 those for 4.50. This seems to be the most sensible way to preserve the
2027 historical information.
2033 1. Minor wording change to the doc/README.SIEVE file.
2035 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2036 computation of the current number of files was incorrect.
2038 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2039 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2042 4. Give more explanation in the error message when the command for a transport
2043 filter fails to execute.
2045 5. There are several places where Exim runs a non-Exim command in a
2046 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2047 was being done only for the command run by the queryprogram router. It is
2048 now done for all such subprocesses. The other cases are: ${run, transport
2049 filters, and the commands run by the lmtp and pipe transports.
2051 6. Added CONFIGURE_GROUP build-time option.
2053 7. Some older OS have a limit of 256 on the maximum number of file
2054 descriptors. Exim was using setrlimit() to set 1000 as a large value
2055 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2056 systems. I've change it so that if it can't get 1000, it tries for 256.
2058 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2059 was an oversight, and furthermore, ever since the addition of extra
2060 controls (e.g. 4.43/32), the checks on when to allow different forms of
2061 "control" were broken. There should now be diagnostics for all cases when a
2062 control that does not make sense is encountered.
2064 9. Added the /retain_sender option to "control=submission".
2066 10. $recipients is now available in the predata ACL (oversight).
2068 11. Tidy the search cache before the fork to do a delivery from a message
2069 received from the command line. Otherwise the child will trigger a lookup
2070 failure and thereby defer the delivery if it tries to use (for example) a
2071 cached ldap connection that the parent has called unbind on.
2073 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2074 of $address_data from the recipient verification was clobbered by the
2075 sender verification.
2077 13. The value of address_data from a sender verification is now available in
2078 $sender_address_data in subsequent conditions in the ACL statement.
2080 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2082 15. Added a new option "connect=<time>" to callout options, to set a different
2085 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2086 was its contents. (It was OK if the option was not defined at all.)
2088 17. A "Completed" log line is now written for messages that are removed from
2089 the spool by the -Mrm option.
2091 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2092 information about exactly what failed.
2094 19. Added -dd to debug only the daemon process.
2096 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2097 handles timeouts, both on the server side and network timeouts. Renamed the
2098 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2100 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2101 It is now set to "smtps".
2103 22. $host_address is now set to the target address during the checking of
2104 ignore_target_hosts.
2106 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2107 being passed; this would have caused $sender_host_name to have been used if
2108 matching the list had actually called for a host name (not very likely,
2109 since this list is usually IP addresses). A host name is now passed as
2112 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2113 code that specifies a non-restarting handler (typically sigaction() in
2114 modern systems) in an attempt to fix a rare and obscure crash bug.
2116 25. Narrowed the window for a race in the daemon that could cause it to ignore
2117 SIGCHLD signals. This is not a major problem, because they are used only to
2118 wake it up if nothing else does.
2120 26. A malformed maildirsize file could cause Exim to calculate negative values
2121 for the mailbox size or file count. Odd effects could occur as a result.
2122 The maildirsize information is now recalculated if the size or filecount
2125 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2126 support for a long time. Removed HAVE_SYS_VFS_H.
2128 28. Installed the latest version of exipick from John Jetmore.
2130 29. In an address list, if the pattern was not a regular expression, an empty
2131 subject address (from a bounce message) matched only if the pattern was an
2132 empty string. Non-empty patterns were not even tested. This was the wrong
2133 because it is perfectly reasonable to use an empty address as part of a
2134 database query. An empty address is now tested by patterns that are
2135 lookups. However, all the other forms of pattern expect the subject to
2136 contain a local part and a domain, and therefore, for them, an empty
2137 address still always fails if the pattern is not itself empty.
2139 30. Exim went into a mad DNS loop when attempting to do a callout where the
2140 host was specified on an smtp transport, and looking it up yielded more
2141 than one IP address.
2143 31. Re-factored the code for checking spool and log partition space into a
2144 function that finds that data and another that does the check. The former
2145 is then used to implement four new variables: $spool_space, $log_space,
2146 $spool_inodes, and $log_inodes.
2148 32. The RFC2047 encoding function was originally intended for short strings
2149 such as real names; it was not keeping to the 75-character limit for
2150 encoded words that the RFC imposes. It now respects the limit, and
2151 generates multiple encoded words if necessary. To be on the safe side, I
2152 have increased the buffer size for the ${rfc2047: expansion operator from
2155 33. It is now permitted to omit both strings after an "if" condition; if the
2156 condition is true, the result is "true". As before, when the second string
2157 is omitted, a false condition yields an empty string. This makes it less
2158 cumbersome to write custom ACL and router conditions.
2160 34. Failure to deliver a bounce message always caused it to be frozen, even if
2161 there was an errors_to setting on the router. The errors_to setting is now
2164 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2165 canonical form (fully expanded) before being placed in
2166 $sender_host_address.
2168 36. The table in the code that translates DNS record types into text (T_A to
2169 "A" for instance) was missing entries for NS and CNAME. It is just possible
2170 that this could have caused confusion if both these types were looked up
2171 for the same domain, because the text type is used as part of Exim's
2172 per-process caching. But the chance of anyone hitting this buglet seems
2175 37. The dnsdb lookup has been extended in a number of ways.
2177 (1) There is a new type, "zns", which walks up the domain tree until it
2178 finds some nameserver records. It should be used with care.
2180 (2) There is a new type, "mxh", which is like "mx" except that it returns
2181 just the host names, not the priorities.
2183 (3) It is now possible to give a list of domains (or IP addresses) to be
2184 looked up. The behaviour when one of the lookups defers can be
2185 controlled by a keyword.
2187 (4) It is now possible to specify the separator character for use when
2188 multiple records are returned.
2190 38. The dnslists ACL condition has been extended: it is now possible to supply
2191 a list of IP addresses and/or domains to be looked up in a particular DNS
2194 39. Added log_selector=+queue_time_overall.
2196 40. When running the queue in the test harness, wait just a tad after forking a
2197 delivery process, to get repeatability of debugging output.
2199 41. Include certificate and key file names in error message when GnuTLS fails
2200 to set them up, because the GnuTLS error message doesn't include the name
2201 of the failing file when there is a problem reading it.
2203 42. Allow both -bf and -bF in the same test run.
2205 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2207 44. The "Exiscan patch" is now merged into the mainline Exim source.
2209 45. Sometimes the final signoff response after QUIT could fail to get
2210 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2211 < 0 before doing a fflush(). This bug looks as though it goes back to the
2212 introduction of TLS in release 3.20, but "sometimes" must have been rare
2213 because the tests only now provoked it.
2215 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2216 (this can affect the format of dates).
2218 47. exim_tidydb, when checking for the continued existence of a message for
2219 which it has found a message-specific retry record, was not finding
2220 messages that were in split spool directories. Consequently, it was
2221 deleting retry records that should have stayed in existence.
2223 48. Steve fixed some bugs in eximstats.
2225 49. The SPA authentication driver was not abandoning authentication and moving
2226 on to the next authenticator when an expansion was forced to fail,
2227 contradicting the general specification for all authenticators. Instead it
2228 was generating a temporary error. It now behaves as specified.
2230 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2231 (the order specifies the preference for clients). The order is now AES256,
2232 AES128, 3DES, ARCFOUR128.
2234 51. Small patch to Sieve code - explicitly set From: when generating an
2237 52. Exim crashed if a remote delivery caused a very long error message to be
2238 recorded - for instance if somebody sent an entire SpamAssassin report back
2239 as a large number of 550 error lines. This bug was coincidentally fixed by
2240 increasing the size of one of Exim's internal buffers (big_buffer) that
2241 happened as part of the Exiscan merge. However, to be on the safe side, I
2242 have made the code more robust (and fixed the comments that describe what
2245 53. Now that there can be additional text after "Completed" in log lines (if
2246 the queue_time_overall log selector is set), a one-byte patch to exigrep
2247 was needed to allow it to recognize "Completed" as not the last thing in
2250 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2251 patch that reportedly fixes this has been added. I am not expert enough to
2252 create a test for it. This is what the patch creator wrote:
2254 "I found a little strange behaviour of ldap code when working with
2255 Windows 2003 AD Domain, where users was placed in more than one
2256 Organization Units. When I tried to give exim partial DN, the exit code
2257 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2258 But simultaneously result of request was absolutely normal ldap result,
2259 so I produce this patch..."
2261 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2262 REFERENCE, so I have modified the code to exclude the patch when that macro
2265 55. Some experimental protocols are using DNS PTR records for new purposes. The
2266 keys for these records are domain names, not reversed IP addresses. The
2267 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2268 leaves it alone. Component reversal etc. now happens only for IP addresses.
2271 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2273 57. Double the size of the debug message buffer (to 2048) so that more of very
2274 long debug lines gets shown.
2276 58. The exicyclog utility now does better if the number of log files to keep
2277 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2279 59. Two changes related to the smtp_active_hostname option:
2281 (1) $smtp_active_hostname is now available as a variable.
2282 (2) The default for smtp_banner uses $smtp_active_hostname instead
2283 of $primary_hostname.
2285 60. The host_aton() function is supposed to be passed a string that is known
2286 to be a valid IP address. However, in the case of IPv6 addresses, it was
2287 not checking this. This is a hostage to fortune. Exim now panics and dies
2288 if the condition is not met. A case was found where this could be provoked
2289 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2290 components; fortuitously, this particular loophole had already been fixed
2291 by change 4.50/55 above.
2293 If there are any other similar loopholes, the new check in host_aton()
2294 itself should stop them being exploited. The report I received stated that
2295 data on the command line could provoke the exploit when Exim was running as
2296 exim, but did not say which command line option was involved. All I could
2297 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2298 running as the user.
2301 61. There was a buffer overflow vulnerability in the SPA authentication code
2302 (which came originally from the Samba project). I have added a test to the
2303 spa_base64_to_bits() function which I hope fixes it.
2306 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2307 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2309 63. The daemon start-up calls getloadavg() while still root for those OS that
2310 need the first call to be done as root, but it missed one case: when
2311 deliver_queue_load_max is set with deliver_drop_privilege. This is
2312 necessary for the benefit of the queue runner, because there is no re-exec
2313 when deliver_drop_privilege is set.
2315 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2316 This has been fixed.
2318 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2319 was in use, was not putting the data itself into the right store pool;
2320 consequently, it could be overwritten for a subsequent message in the same
2321 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2324 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2326 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2327 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2328 the function were treating the return as a boolean value, which happened to
2329 work because 0=false and not-0=true, but is not correct code.
2331 68. The host_aton() function was not handling scoped IPv6 addresses (those
2332 with, for example, "%eth0" on the end) correctly.
2334 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2335 negated items (that is, ~something) in unsigned ints. Some compilers
2336 apparently mutter when there is no cast.
2338 70. If an address verification called from an ACL failed, and did not produce a
2339 user-specific message (i.e. there was only a "system" message), nothing was
2340 put in $acl_verify_message. In this situation, it now puts the system
2343 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2344 session; change 4.31/43 added the unwanted input to the log line - except
2345 that it did not do this in the start of session case. It now does.
2347 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2348 This is wrong and can cause the other end to generate a synchronization
2349 error if it is another Exim or anything else that does the synchronization
2350 check. A QUIT command is no longer sent after a timeout.
2352 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2353 during host lookups.
2355 74. The defer_ok option of callout verification was not working if it was used
2356 when verifying addresses in header lines, that is, for this case:
2358 verify = header_sender/callout=defer_ok
2360 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2361 those file descriptors could be used for SMTP connections. If anything
2362 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2363 could be sent to the SMTP client, causing chaos. The daemon now opens
2364 stdin, stdout, and stderr to /dev/null when it puts itself into the
2367 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2368 log by default. The user can override this with suitable Perl magic.
2370 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2371 the log message when discard triggers, was not working for the DATA ACL or
2372 for the non-SMTP ACL.
2374 78. Error message wording change in sieve.c.
2376 79. If smtp_accept_max_per_host was set, the number of connections could be
2377 restricted to fewer than expected, because the daemon was trying to set up
2378 a new connection before checking whether the processes handling previous
2379 connections had finished. The check for completed processes is now done
2380 earlier. On busy systems, this bug wouldn't be noticed because something
2381 else would have woken the daemon, and it would have reaped the completed
2384 80. If a message was submitted locally by a user whose login name contained one
2385 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2386 It caused a spool format error. I have fixed the spool reading code. A
2387 related problem was that the "from" clause in the Received: line became
2388 illegal because of the space(s). It is now covered by ${quote_local_part.
2390 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2393 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2395 83. Patch from Sieve maintainer to fix unterminated string problem in
2396 "vacation" handling.
2398 84. Some minor changes to the Linux configuration files to help with other
2399 OS variants using glibc.
2401 85. One more patch for Sieve to update vacation handling to latest spec.
2404 ----------------------------------------------------
2405 See the note above about the 4.44 and 4.50 releases.
2406 ----------------------------------------------------
2412 1. Change 4.43/35 introduced a bug that caused file counts to be
2413 incorrectly computed when quota_filecount was set in an appendfile
2416 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2417 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2420 3. Give more explanation in the error message when the command for a transport
2421 filter fails to execute.
2423 4. There are several places where Exim runs a non-Exim command in a
2424 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2425 was being done only for the command run by the queryprogram router. It is
2426 now done for all such subprocesses. The other cases are: ${run, transport
2427 filters, and the commands run by the lmtp and pipe transports.
2429 5. Some older OS have a limit of 256 on the maximum number of file
2430 descriptors. Exim was using setrlimit() to set 1000 as a large value
2431 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2432 systems. I've change it so that if it can't get 1000, it tries for 256.
2434 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2435 was an oversight, and furthermore, ever since the addition of extra
2436 controls (e.g. 4.43/32), the checks on when to allow different forms of
2437 "control" were broken. There should now be diagnostics for all cases when a
2438 control that does not make sense is encountered.
2440 7. $recipients is now available in the predata ACL (oversight).
2442 8. Tidy the search cache before the fork to do a delivery from a message
2443 received from the command line. Otherwise the child will trigger a lookup
2444 failure and thereby defer the delivery if it tries to use (for example) a
2445 cached ldap connection that the parent has called unbind on.
2447 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2448 of $address_data from the recipient verification was clobbered by the
2449 sender verification.
2451 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2452 was its contents. (It was OK if the option was not defined at all.)
2454 11. A "Completed" log line is now written for messages that are removed from
2455 the spool by the -Mrm option.
2457 12. $host_address is now set to the target address during the checking of
2458 ignore_target_hosts.
2460 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2461 being passed; this would have caused $sender_host_name to have been used if
2462 matching the list had actually called for a host name (not very likely,
2463 since this list is usually IP addresses). A host name is now passed as
2466 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2467 code that specifies a non-restarting handler (typically sigaction() in
2468 modern systems) in an attempt to fix a rare and obscure crash bug.
2470 15. Narrowed the window for a race in the daemon that could cause it to ignore
2471 SIGCHLD signals. This is not a major problem, because they are used only to
2472 wake it up if nothing else does.
2474 16. A malformed maildirsize file could cause Exim to calculate negative values
2475 for the mailbox size or file count. Odd effects could occur as a result.
2476 The maildirsize information is now recalculated if the size or filecount
2479 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2480 support for a long time. Removed HAVE_SYS_VFS_H.
2482 18. Updated exipick to current release from John Jetmore.
2484 19. Allow an empty sender to be matched against a lookup in an address list.
2485 Previously the only cases considered were a regular expression, or an
2488 20. Exim went into a mad DNS lookup loop when doing a callout where the
2489 host was specified on the transport, if the DNS lookup yielded more than
2492 21. The RFC2047 encoding function was originally intended for short strings
2493 such as real names; it was not keeping to the 75-character limit for
2494 encoded words that the RFC imposes. It now respects the limit, and
2495 generates multiple encoded words if necessary. To be on the safe side, I
2496 have increased the buffer size for the ${rfc2047: expansion operator from
2499 22. Failure to deliver a bounce message always caused it to be frozen, even if
2500 there was an errors_to setting on the router. The errors_to setting is now
2503 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2504 canonical form (fully expanded) before being placed in
2505 $sender_host_address.
2507 24. Updated eximstats to version 1.33
2509 25. Include certificate and key file names in error message when GnuTLS fails
2510 to set them up, because the GnuTLS error message doesn't include the name
2511 of the failing file when there is a problem reading it.
2513 26. Expand error message when OpenSSL has problems setting up cert/key files.
2516 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2517 (this can affect the format of dates).
2519 28. exim_tidydb, when checking for the continued existence of a message for
2520 which it has found a message-specific retry record, was not finding
2521 messages that were in split spool directories. Consequently, it was
2522 deleting retry records that should have stayed in existence.
2524 29. eximstats updated to version 1.35
2525 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2526 1.35 - bugfix such that pie charts by volume are generated correctly
2528 30. The SPA authentication driver was not abandoning authentication and moving
2529 on to the next authenticator when an expansion was forced to fail,
2530 contradicting the general specification for all authenticators. Instead it
2531 was generating a temporary error. It now behaves as specified.
2533 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2534 (the order specifies the preference for clients). The order is now AES256,
2535 AES128, 3DES, ARCFOUR128.
2537 31. Small patch to Sieve code - explicitly set From: when generating an
2540 32. Exim crashed if a remote delivery caused a very long error message to be
2541 recorded - for instance if somebody sent an entire SpamAssassin report back
2542 as a large number of 550 error lines. This bug was coincidentally fixed by
2543 increasing the size of one of Exim's internal buffers (big_buffer) that
2544 happened as part of the Exiscan merge. However, to be on the safe side, I
2545 have made the code more robust (and fixed the comments that describe what
2548 33. Some experimental protocols are using DNS PTR records for new purposes. The
2549 keys for these records are domain names, not reversed IP addresses. The
2550 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2551 leaves it alone. Component reversal etc. now happens only for IP addresses.
2554 34. The host_aton() function is supposed to be passed a string that is known
2555 to be a valid IP address. However, in the case of IPv6 addresses, it was
2556 not checking this. This is a hostage to fortune. Exim now panics and dies
2557 if the condition is not met. A case was found where this could be provoked
2558 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2559 components; fortuitously, this particular loophole had already been fixed
2560 by change 4.50/55 or 4.44/33 above.
2562 If there are any other similar loopholes, the new check in host_aton()
2563 itself should stop them being exploited. The report I received stated that
2564 data on the command line could provoke the exploit when Exim was running as
2565 exim, but did not say which command line option was involved. All I could
2566 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2567 running as the user.
2570 35. There was a buffer overflow vulnerability in the SPA authentication code
2571 (which came originally from the Samba project). I have added a test to the
2572 spa_base64_to_bits() function which I hope fixes it.
2575 36. The daemon start-up calls getloadavg() while still root for those OS that
2576 need the first call to be done as root, but it missed one case: when
2577 deliver_queue_load_max is set with deliver_drop_privilege. This is
2578 necessary for the benefit of the queue runner, because there is no re-exec
2579 when deliver_drop_privilege is set.
2581 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2582 was in use, was not putting the data itself into the right store pool;
2583 consequently, it could be overwritten for a subsequent message in the same
2584 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2587 38. Sometimes the final signoff response after QUIT could fail to get
2588 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2589 < 0 before doing a fflush(). This bug looks as though it goes back to the
2590 introduction of TLS in release 3.20, but "sometimes" must have been rare
2591 because the tests only now provoked it.
2597 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2598 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2599 Now it returns an integer. A number of calls were still expecting a T/F
2600 return. Fortuitously, in all cases, the tests worked in OK situations,
2601 which is the norm. However, things would have gone wrong on any write
2602 failures on the smtp file descriptor. This function is used when sending
2603 messages over SMTP and also when doing verify callouts.
2605 2. When Exim is called to do synchronous delivery of a locally submitted
2606 message (the -odf or -odi options), it no longer closes stderr before doing
2609 3. Implemented the mua_wrapper option.
2611 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2613 5. Implemented the functions header_remove(), header_testname(),
2614 header_add_at_position(), and receive_remove_recipient(), and exported them
2617 6. If an ACL "warn" statement specified the addition of headers, Exim already
2618 inserted X-ACL-Warn: at the start if there was no header name. However, it
2619 was not making this test for the second and subsequent header lines if
2620 there were newlines in the string. This meant that an invalid header could
2621 be inserted if Exim was badly configured.
2623 7. Allow an ACL "warn" statement to add header lines at the start or after all
2624 the Received: headers, as well as at the end.
2626 8. Added the rcpt_4xx retry error code.
2628 9. Added postmaster_mailfrom=xxx to callout verification option.
2630 10. Added mailfrom=xxxx to the callout verification option, for verify=
2633 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2634 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2636 12. Inserted some casts to stop certain compilers warning when using pointer
2637 differences as field lengths or precisions in printf-type calls (mostly
2638 affecting debugging statements).
2640 13. Added optional readline() support for -be (dynamically loaded).
2642 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2643 same clock tick as a message's arrival, so that its received time was the
2644 same as the "first fail" time on the retry record, and that message
2645 remained on the queue past the ultimate address timeout, every queue runner
2646 would try a delivery (because it was past the ultimate address timeout) but
2647 after another failure, the ultimate address timeout, which should have then
2648 bounced the address, did not kick in. This was a "< instead of <=" error;
2649 in most cases the first failure would have been in the next clock tick
2650 after the received time, and all would be well.
2652 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2653 being recognized when the domain list was tested by the match_domain
2654 condition in an expansion string.
2656 16. Added the ${str2b64: operator.
2658 17. Exim was always calling setrlimit() to set a large limit for the number of
2659 processes, without checking whether the existing limit was already
2660 adequate. (It did check for the limit on file descriptors.) Furthermore,
2661 errors from getrlimit() and setrlimit() were being ignored. Now they are
2662 logged to the main and panic logs, but Exim does carry on, to try to do its
2663 job under whatever limits there are.
2665 18. Imported PCRE 5.0.
2667 19. Trivial typo in log message " temporarily refused connection" (the leading
2670 20. If the log selector return_path_on_delivery was set and an address was
2671 redirected to /dev/null, the delivery process crashed because it assumed
2672 that a return path would always be set for a "successful" delivery. In this
2673 case, the whole delivery is bypassed as an optimization, and therefore no
2676 21. Internal re-arrangement: the function for sending a challenge and reading
2677 a response while authentication was assuming a zero-terminated challenge
2678 string. It's now changed to take a pointer and a length, to allow for
2679 binary data in such strings.
2681 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2683 23. Exim was not respecting finduser_retries when seeking the login of the
2684 uid under which it was called; it was always trying 10 times. (The default
2685 setting of finduser_retries is zero.) Also, it was sleeping after the final
2686 failure, which is pointless.
2688 24. Implemented tls_on_connect_ports.
2690 25. Implemented acl_smtp_predata.
2692 26. If the domain in control=submission is set empty, Exim assumes that the
2693 authenticated id is a complete email address when it generates From: or
2694 Sender: header lines.
2696 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2697 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2698 chown and chgrp in /bin and hostname in /usr/bin.
2700 28. Exim was keeping the "process log" file open after each use, just as it
2701 does for the main log. This opens the possibility of it remaining open for
2702 long periods when the USR1 signal hits a daemon. Occasional processlog
2703 errors were reported, that could have been caused by this. Anyway, it seems
2704 much more sensible not to leave this file open at all, so that is what now
2707 29. The long-running daemon process does not normally write to the log once it
2708 has entered its main loop, and it closes the log before doing so. This is
2709 so that log files can straightforwardly be renamed and moved. However,
2710 there are a couple of unusual error situations where the daemon does write
2711 log entries, and I had neglected to close the log afterwards.
2713 30. The text of an SMTP error response that was received during a remote
2714 delivery was being truncated at 512 bytes. This is too short for some of
2715 the long messages that one sometimes sees. I've increased the limit to
2718 31. It is now possible to make retry rules that apply only when a message has a
2719 specific sender, in particular, an empty sender.
2721 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2722 it possible to be selective about when SMTP synchronization is enforced.
2724 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2726 32. Implemented hosts_connection_nolog.
2728 33. Added an ACL for QUIT.
2730 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2733 35. Added mailbox_size and mailbox_filecount to appendfile.
2735 36. Added control = no_multiline_responses to ACLs.
2737 37. There was a bug in the logic of the code that waits for the clock to tick
2738 in the case where the clock went backwards by a substantial amount such
2739 that the microsecond fraction of "now" was more than the microsecond
2740 fraction of "then" (but the whole seconds number was less).
2742 38. Added support for the libradius Radius client library this is found on
2743 FreeBSD (previously only the radiusclient library was supported).
2749 1. When certain lookups returned multiple values in the form name=value, the
2750 quoting of the values was not always being done properly. Specifically:
2751 (a) If the value started with a double quote, but contained no whitespace,
2753 (b) If the value contained whitespace other than a space character (i.e.
2754 tabs or newlines or carriage returns) it was not quoted.
2755 This fix has been applied to the mysql and pgsql lookups by writing a
2756 separate quoting function and calling it from the lookup code. The fix
2757 should probably also be applied to nisplus, ibase and oracle lookups, but
2758 since I cannot test any of those, I have not disturbed their existing code.
2760 2. A hit in the callout cache for a specific address caused a log line with no
2761 reason for rejecting RCPT. Now it says "Previous (cached) callout
2762 verification failure".
2764 3. There was an off-by-one bug in the queryprogram router. An over-long
2765 return line was truncated at 256 instead of 255 characters, thereby
2766 overflowing its buffer with the terminating zero. As well as fixing this, I
2767 have increased the buffer size to 1024 (and made a note to document this).
2769 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2770 when Exim is waiting for an SMTP response from a remote server, Exim
2771 restarts its select() call on the socket, thereby resetting its timeout.
2772 This is not a problem when such interrupts are rare. Somebody set up a cron
2773 job to run exiwhat every 2 minutes, which is less than the normal select()
2774 timeout (5 or 10 minutes). This meant that the select() timeout never
2775 kicked in because it was always reset. I have fixed this by comparing the
2776 time when an interrupt arrives with the time at the start of the first call
2777 to select(). If more time than the timeout has elapsed, the interrupt is
2778 treated as a timeout.
2780 5. Some internal re-factoring in preparation for the addition of Sieve
2781 extensions (by MH). In particular, the "personal" test is moved to a
2782 separate function, and given an option for scanning Cc: and Bcc: (which is
2783 not set for Exim filters).
2785 6. When Exim created an email address using the login of the caller as the
2786 local part (e.g. when creating a From: or Sender: header line), it was not
2787 quoting the local part when it contained special characters such as @.
2789 7. Installed new OpenBSD configuration files.
2791 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2792 try to make them clearer.
2794 9. Callout options, other than the timeout value, were being ignored when
2795 verifying sender addresses in header lines. For example, when using
2797 verify = header_sender/callout=no_cache
2799 the cache was (incorrectly) being used.
2801 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2802 only the Cygwin environment.
2804 11. When return_path_on_delivery was set as a log selector, if different remote
2805 addresses in the same message used different return paths and parallel
2806 remote delivery occurred, the wrong values would sometimes be logged.
2807 (Whenever a remote delivery process finished, the return path value from
2808 the most recently started remote delivery process was logged.)
2810 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2811 lines when AUTH and/or TLS are in use. This is the "received protocol"
2812 field. Exim used to use "asmtp" for authenticated SMTP, without any
2813 indication (in the protocol name) for TLS use. Now it follows the RFC and
2814 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2815 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2816 names appear in log lines as well as in Received: header lines.
2818 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2819 extensions, and comparison tests, and to fix some bugs.
2821 14. Changes to the "personal" filter test:
2823 (1) The test was buggy in that it was just doing the equivalent of
2824 "contains" tests on header lines. For example, if a user's address was
2825 anne@some.where, the "personal" test would incorrectly be true for
2827 To: susanne@some.where
2829 This test is now done by extracting each address from the header in turn,
2830 and checking the entire address. Other tests that are part of "personal"
2831 are now done using regular expressions (for example, to check local parts
2832 of addresses in From: header lines).
2834 (2) The list of non-personal local parts in From: addresses has been
2835 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2836 taken from the Sieve specification recommendations.
2838 (3) If the message contains any header line starting with "List-" it is
2839 treated as non-personal.
2841 (4) The test for "circular" in the Subject: header line has been removed
2842 because it now seems ill-conceived.
2844 15. Minor typos in src/EDITME comments corrected.
2846 16. Installed latest exipick from John Jetmore.
2848 17. If headers_add on a router specified a text string that was too long for
2849 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2850 of string_sprintf() is now avoided.
2852 18. $message_body_size was not set (it was always zero) when running the DATA
2853 ACL and the local_scan() function.
2855 19. For the "mail" command in an Exim filter, no default was being set for
2856 the once_repeat time, causing a random time value to be used if "once" was
2857 specified. (If the value happened to be <= 0, no repeat happened.) The
2858 default is now 0s, meaning "never repeat". The "vacation" command was OK
2859 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2860 (I found it when inspecting the code).
2862 20. There is now an overall timeout for performing a callout verification. It
2863 defaults to 4 times the callout timeout, which applies to individual SMTP
2864 commands during the callout. The overall timeout applies when there is more
2865 than one host that can be tried. The timeout is checked before trying the
2866 next host. This prevents very long delays if there are a large number of
2867 hosts and all are timing out (e.g. when the network connections are timing
2868 out). The value of the overall timeout can be changed by specifying an
2869 additional sub-option for "callout", called "maxwait". For example:
2871 verify = sender/callout=5s,maxwait=20s
2873 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2874 to the end before writing, but this should make it even safer).
2876 22. Exim was forgetting that it had advertised PIPELINING for the second and
2877 subsequent messages on an SMTP connection. It was also not resetting its
2878 memory on STARTTLS and an internal HELO.
2880 23. When Exim logs an SMTP synchronization error within a session, it now
2881 records whether PIPELINING has been advertised or not.
2883 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2884 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2885 rather than long int.
2887 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2889 26. Added the never_mail option to autoreply.
2895 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2896 crash if the getsockname() call failed; this can happen if a connection is
2897 closed very soon after it is established. The problem was simply in the
2898 order in which certain operations were done, causing Exim to try to write
2899 to the SMTP stream before it had set up the file descriptor. The bug has
2900 been fixed by making things happen in the correct order.
2906 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2907 before the connection was closed, thus losing the rejection response.
2909 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2910 some early Solaris releases, but causes trouble in current releases where
2911 socklen_t is defined.
2913 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2916 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2919 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2920 that releasing the top of it at the end releases what was used for sub-
2921 expansions (unless the block got too big). However, discard this block if
2922 the first thing is a variable or header, so that we can use its block when
2923 it is dynamic (useful for very large $message_headers, for example).
2925 6. Lookups now cache *every* query, not just the most recent. A new, separate
2926 store pool is used for this. It can be recovered when all lookup caches are
2927 flushed. Lookups now release memory at the end of their result strings.
2928 This has involved some general refactoring of the lookup sources.
2930 7. Some code has been added to the store_xxx() functions to reduce the amount
2931 of flapping under certain conditions.
2933 8. log_incoming_interface used to affect only the <= reception log lines. Now
2934 it causes the local interface and port to be added to several more SMTP log
2935 lines, for example "SMTP connection from", and rejection lines.
2937 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2939 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2941 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2942 could be overwritten at the end of the current message (or the start of a
2943 new message if it was set in a HELO ACL). The value is now preserved for
2944 the duration of the SMTP connection.
2946 12. If a transport had a headers_rewrite setting, and a matching header line
2947 contained an unqualified address, that address was qualified, even if it
2948 did not match any rewriting rules. The underlying bug was that the values
2949 of the flags that permit the existence of unqualified sender and recipient
2950 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2951 non-local messages, and by -bnq for local messages) were not being
2952 preserved with the message after it was received.
2954 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2955 "next input=" as part of the text comprising the host identity instead of
2956 the correct text. The code was using the same buffer for two different
2957 strings. However, depending on which order the printing function evaluated
2958 its arguments, the bug did not always show up. Under Linux, for example, my
2959 test suite worked just fine.
2961 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2962 doesn't work with some older versions of Perl. It has been changed to "my",
2963 which in any case is probably the better facility to use.
2965 15. A really picky compiler found some instances of statements for creating
2966 error messages that either had too many or two few arguments for the format
2969 16. The size of the buffer for calls to the DNS resolver has been increased
2970 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2971 for addresses that have a lot of PTR records. This alleviates a problem; it
2972 does not fully solve it.
2974 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2975 buffer now truncates the list and logs the incident, which is the same
2976 action as happens when Exim is looking up a host name and its aliases.
2977 Previously in this situation something unpredictable would happen;
2978 sometimes it was "internal error: store_reset failed".
2980 18. If a server dropped the connection unexpectedly when an Exim client was
2981 using GnuTLS and trying to read a response, the client delivery process
2982 crashed while trying to generate an error log message.
2984 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2985 string, for example:
2987 warn message = H1: something\nH2: something
2989 the text was added as a single header line from Exim's point of view
2990 though it ended up OK in the delivered message. However, searching for the
2991 second and subsequent header lines using $h_h2: did not work. This has been
2992 fixed. Similarly, if a system filter added multiple headers in this way,
2993 the routers could not see them.
2995 20. Expanded the error message when iplsearch is called with an invalid key to
2996 suggest using net-iplsearch in a host list.
2998 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2999 ACLs are no longer actually imposed (and a message to that effect is
3002 22. If a "gecos" field in a passwd entry contained escaped characters, in
3003 particular, if it contained a \" sequence, Exim got it wrong when building
3004 a From: or a Sender: header from that name. A second bug also caused
3005 incorrect handling when an unquoted " was present following a character
3006 that needed quoting.
3008 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3009 was not being matched caselessly.
3011 24. Arranged for all hyphens in the exim.8 source to be escaped with
3014 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3015 properly. Recipient callout cache records were still being keyed to include
3016 the sender, even when use_sender was set false. This led to far more
3017 callouts that were necessary. The sender is no longer included in the key
3018 when use_sender is false.
3020 26. Added "control = submission" modifier to ACLs.
3022 27. Added the ${base62d: operator to decode base 62 numbers.
3024 28. dnsdb lookups can now access SRV records.
3026 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3027 the configuration file.
3029 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3030 (-v) mode. This makes the output for a verbose queue run more intelligible.
3032 31. Added a use_postmaster feature to recipient callouts.
3034 32. Added the $body_zerocount variable, containing the number of binary zero
3035 bytes in the message body.
3037 33. The time of last modification of the "new" subdirectory is now used as the
3038 "mailbox time last read" when there is a quota error for a maildir
3041 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3043 35. Added +ignore_unknown as a special item in host lists.
3045 36. Code for decoding IPv6 addresses in host lists is now included, even if
3046 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3047 address was recognized as an IP address, but was then not correctly decoded
3048 into binary, causing unexpected and incorrect effects when compared with
3055 1. Very minor rewording of debugging text in manualroute to say "list of
3056 hosts" instead of "hostlist".
3058 2. If verify=header_syntax was set, and a header line with an unqualified
3059 address (no domain) and a large number of spaces between the end of the
3060 name and the colon was received, the reception process suffered a buffer
3061 overflow, and (when I tested it) crashed. This was caused by some obsolete
3062 code that should have been removed. The fix is to remove it!
3064 3. When running in the test harness, delay a bit after writing a bounce
3065 message to get a bit more predictability in the log output.
3067 4. Added a call to search_tidyup() just before forking a reception process. In
3068 theory, someone could use a lookup in the expansion of smtp_accept_max_
3069 per_host which, without the tidyup, could leave open a database connection.
3071 5. Added the variables $recipient_data and $sender_data which get set from a
3072 lookup success in an ACL "recipients" or "senders" condition, or a router
3073 "senders" option, similar to $domain_data and $local_part_data.
3075 6. Moved the writing of debug_print from before to after the "senders" test
3078 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3079 problems for message scanning, either using a data ACL, or using
3080 local_scan() because the Received: header was not generated till after they
3081 were called (in order to set the time as the time of reception completion).
3082 I have revised the way this works. The header is now generated after the
3083 body is received, but before the ACL or local_scan() are called. After they
3084 are run, the timestamp in the header is updated.
3090 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3091 before starting a queue runner without re-exec. This happened only when
3092 deliver_drop_privilege was set or when the Exim user was set to root. The
3093 effect of the bug was that timeouts during subsequent deliveries caused
3094 crashes instead of being properly handled. The handler is now left at its
3095 default (and expected) setting.
3097 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3098 message, again when deliver_drop_privilege is set or Exim is run as root.
3099 The bug described in (1) was not present in this case, but the tidying up
3100 of the other signals was missing. I have made the two cases consistent.
3102 3. The ignore_target_hosts setting on a manualroute router was being ignored
3103 for hosts that were looked up using the /MX notation.
3105 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3108 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3109 operated on the sender address. After changing the $sender_address to <>
3110 for the sender address verify, Exim was re-instated it as the original
3111 (before rewriting) address, but remembering that it had rewritten it, so it
3112 wasn't rewriting it again. This bug also had the effect of breaking the
3113 sender address verification caching when the sender address was rewritten.
3115 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3116 This has been changed so that if the ip literal address matches
3117 ignore_target_hosts, the router declines.
3119 7. Added expansion conditions match_domain, match_address, and match_local_
3120 part (NOT match_host).
3122 8. The placeholder for the Received: header didn't have a length field set.
3124 9. Added code to Exim itself and to exim_lock to test for a specific race
3125 condition that could lead to file corruption when using MBX delivery. The
3126 issue is with the lockfile that is created in /tmp. If this file is removed
3127 after a process has opened it but before that process has acquired a lock,
3128 there is the potential for a second process to recreate the file and also
3129 acquire a lock. This could lead to two Exim processes writing to the file
3130 at the same time. The added code performs the same test as UW imapd; it
3131 checks after acquiring the lock that its file descriptor still refers to
3132 the same named file.
3134 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3135 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3138 11. Added the smtp_active_hostname option. If used, this will typically be made
3139 to depend on the incoming interface address. Because $interface_address is
3140 not set up until the daemon has forked a reception process, error responses
3141 that can happen earlier (such as "too many connections") no longer contain
3144 12. If an expansion in a condition on a "warn" statement fails because a lookup
3145 defers, the "warn" statement is abandoned, and the next ACL statement is
3146 processed. Previously this caused the whole ACL to be aborted.
3148 13. Added the iplsearch lookup type.
3150 14. Added ident_timeout as a log selector.
3152 15. Added tls_certificate_verified as a log selector.
3154 16. Added a global option tls_require_ciphers (compare the smtp transport
3155 option of the same name). This controls incoming TLS connections.
3157 17. I finally figured out how to make tls_require_ciphers do a similar thing
3158 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3159 before starting the TLS session.
3161 18. Tabs are now shown as \t in -bP output.
3163 19. If the log selector return_path_on_delivery was set, Exim crashed when
3164 bouncing a message because it had too many Received: header lines.
3166 20. If two routers both had headers_remove settings, and the first one included
3167 a superfluous trailing colon, the final name in the first list and the
3168 first name in the second list were incorrectly joined into one item (with a
3169 colon in the middle).
3175 1. Added -C and -D options to the exinext utility, mainly to make it easier
3176 to include in the automated testing, but these could be helpful when
3177 multiple configurations are in use.
3179 2. The exinext utility was not formatting the output nicely when there was
3180 an alternate port involved in the retry record key, nor when there was a
3181 message id as well (for retries that were specific to a specific message
3182 and a specific host). It was also confused by IPv6 addresses, because of
3183 the additional colons they contain. I have fixed the IPv4 problem, and
3184 patched it up to do a reasonable job for IPv6.
3186 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3187 delivery, the log line now contains "pipelined" if PIPELINING was used.
3189 4. An SMTP transport process used to panic and die if the bind() call to set
3190 an explicit outgoing interface failed. This has been changed; it is now
3191 treated in the same way as a connect() failure.
3193 5. A reference to $sender_host_name in the part of a conditional expansion
3194 that was being skipped was still causing a DNS lookup. This no longer
3197 6. The def: expansion condition was not recognizing references to header lines
3198 that used bh_ and bheader_.
3200 7. Added the _cache feature to named lists.
3202 8. The code for checking quota_filecount in the appendfile transport was
3203 allowing one more file than it should have been.
3205 9. For compatibility with Sendmail, the command line option
3213 and sets the incoming protocol and host name (for trusted callers). The
3214 host name and its colon can be omitted when only the protocol is to be set.
3215 Note the Exim already has two private options, -pd and -ps, that refer to
3216 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3217 "s", but I don't think that's a major issue.
3219 10. A number of refactoring changes to the code, none of which should affect
3222 (a) The number of logging options was getting close to filling up the
3223 32-bit word that was used as a bit map. I have split them into two classes:
3224 those that are passed in the argument to log_write(), and those that are
3225 only ever tested independently outside of that function. These are now in
3226 separate 32-bit words, so there is plenty of room for expansion again.
3227 There is no change in the user interface or the logging behaviour.
3229 (b) When building, for example, log lines, the code previously used a
3230 macro that called string_cat() twice, in order to add two strings. This is
3231 not really sufficiently general. Furthermore, there was one instance where
3232 it was actually wrong because one of the argument was used twice, and in
3233 one call a function was used. (As it happened, calling the function twice
3234 did not affect the overall behaviour.) The macro has been replaced by a
3235 function that can join an arbitrary number of extra strings onto a growing
3238 (c) The code for expansion conditions now uses a table and a binary chop
3239 instead of a serial search (which was left over from when there were very
3240 few conditions). Also, it now recognizes conditions like "pam" even when
3241 the relevant support is not compiled in: a suitably worded error message is
3242 given if an attempt is made to use such a condition.
3244 11. Added ${time_interval:xxxxx}.
3246 12. A bug was causing one of the ddress fields not to be passed back correctly
3247 from remote delivery subprocesses. The field in question was not being
3248 subsequently used, so this caused to problems in practice.
3250 13. Added new log selectors queue_time and deliver_time.
3252 14. Might have fixed a bug in maildirsizefile handling that threw up
3253 "unexpected character" debug warnings, and recalculated the data
3254 unnecessarily. In any case, I expanded the warning message to give more
3257 15. Added the message "Restricted characters in address" to the statements in
3258 the default ACL that block characters like @ and % in local parts.
3260 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3261 Three changes have been made:
3263 (a) There was a serious bug; a negative response to MAIL caused the whole
3264 recipient domain to be cached as invalid, thereby blocking all messages
3265 to all local parts at the same domain, from all senders. This bug has
3266 been fixed. The domain is no longer cached after a negative response to
3267 MAIL if the sender used is not empty.
3269 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3272 (c) A new callout option, "use_sender" has been added for people who want
3273 the modified behaviour.
3279 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3282 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3283 indeed breaks things for older releases.
3285 3. Added additional logging to the case where there is a problem reading data
3286 from a filter that is running in a subprocess using a pipe, in order to
3287 try to track down a specific problem.
3289 4. Testing facility fudge: when running in the test harness and attempting
3290 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3291 getting "No route to host". Convert this to a timeout.
3293 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3296 6. Some OS don't have socklen_t but use size_t instead. This affects the
3297 fifth argument of getsockopt() amongst other things. This is now
3298 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3299 can be set for individual OS. I have set it for SunOS5, OSF1, and
3300 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3301 some earlier ones do not.
3303 7. Change 4.30/15 was not doing the test caselessly.
3305 8. The standard form for an IPv6 address literal was being rejected by address
3306 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3307 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3308 this, as well as the form without the "IPv6" on the front (but only when
3309 address literals are enabled, of course).
3311 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3313 10. Exim crashed if a message with an empty sender address specified by -f
3314 encountered a router with an errors_to setting. This could be provoked only
3315 by a command such as
3319 where an empty string was supplied; "<>" did not hit this bug.
3321 11. Installed PCRE release 4.5.
3323 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3324 remained set. It is now erased.
3326 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3327 times from message ids (which are base 36 rather than the normal 62).
3329 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3330 were being counted as actual protocol errors, and logged if the log
3331 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3332 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3333 and DATA following a set of rejected RCPTs do not count as protocol errors.
3334 In other words, Exim assumes they were pipelined, though this may not
3335 actually be the case. Of course, in all cases the client gets an
3336 appropriate error code.
3338 15. If a lookup fails in an ACL condition, a message about the failure may
3339 be available; it is used if testing the ACL cannot continue, because most
3340 such messages specify what the cause of the deferral is. However, some
3341 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3342 that caused an old message to be retained and used if a later statement
3343 caused a defer, replacing the real cause of the deferral.
3345 16. If an IP address had so many PTR records that the DNS lookup buffer
3346 was not large enough to hold them, Exim could crash while trying to process
3347 the truncated data. It now detects and logs this case.
3349 17. Further to 4.21/58, another change has been made: if (and only if) the
3350 first line of a message (the first header line) ends with CRLF, a bare LF
3351 in a subsequent header line has a space inserted after it, so as not to
3352 terminate the header.
3354 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3355 unnecessarily, used atoi() instead of strtol(), and didn't check the
3356 termination when getting file sizes from file names by regex.
3358 19. Completely re-implemented the support for maildirsize files, in the light
3359 of a number of problems with the previous contributed implementation
3360 (4.30/29). In particular:
3362 . If the quota is zero, the maildirsize file is maintained, but no quota is
3365 . If the maildir directory does not exist, it is created before any attempt
3366 to write a maildirsize file.
3368 . The quota value in the file is just a cache; if the quota is changed in
3369 the transport, the new value overrides.
3371 . A regular expression is available for excluding directories from the
3374 20. The autoreply transport checks the characters in options that define the
3375 message's headers; it allows continued headers, but it was checking with
3376 isspace() after an embedded newline instead of explicitly looking for a
3379 21. If all the "regular" hosts to which an address was routed had passed their
3380 expiry times, and had not reached their retry times, the address was
3381 bounced, even if fallback hosts were defined. Now Exim should go on to try
3384 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3385 equivalent code in the SMTP transport. Some hosts send humungous responses
3386 to HELO/EHLO, more than 1024 it seems.
3388 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3389 gives compiler warnings in some environments. I've now done it "properly",
3392 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3393 (because of problems with the built-in one) was declared to return uschar *
3394 instead of char *, causing compiler failure.
3396 25. Fixed a file descriptor leak when processing alias/forward files.
3398 26. Fixed a minor format string issue in dbfn.c.
3400 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3402 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3403 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3406 29. When a sender address is verified, it is cached, to save repeating the test
3407 when there is more than one recipient in a message. However, when the
3408 verification involves a callout, it is possible for different callout
3409 options to be set for different recipients. It is too complicated to keep
3410 track of this in the cache, so now Exim always runs a verification when a
3411 callout is required, relying on the callout cache for the optimization.
3412 The overhead is duplication of the address routing, but this should not be
3415 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3416 to be verified with callout=postmaster, and the main callout worked but the
3417 postmaster check failed, the verification correctly failed. However, if a
3418 subsequent RCPT command asked for sender verification *without* the
3419 postmaster check, incorrect caching caused this verification also to fail,
3422 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3423 it was not caching the DNS options (qualify_single, search_parents) that
3424 were used when the lookup failed. A subsequent lookup with different
3425 options therefore always gave the same answer, though there were cases
3426 where it should not have. (Example: a "domains = !$mx_any" option on a
3427 dnslookup router: the "domains" option is always processed without any
3428 widening, but the router might have qualify_single set.) Now Exim uses the
3429 cached value only when the same options are set.
3431 32. Added John Jetmore's "exipick" utility to the distribution.
3433 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3434 than a timeout (e.g. a certificate is required, and is not provided), an
3435 Exim server now closes the connection immediately. Previously it waited for
3436 the client to close - but if the client is SSL, it seems that they each
3437 wait for each other, leading to a delay before one of them times out.
3439 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3440 maintained 0.8.x compatibility because I don't think many are using it, and
3441 it is clearly obsolete.
3443 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3446 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3447 client certificate was expired. A simple patch fixes this, though I don't
3448 understand the full logic of why the verify callback is called multiple
3451 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3452 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3453 which causes problems with some clients (such as the Certicom SSL Plus
3454 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3455 disables the coutermeasure allowing Eudora to connect."
3457 38. Exim was not checking that a write() to a log file succeeded. This could
3458 lead to Bad Things if a log got too big, in particular if it hit a file
3459 size limit. Exim now panics and dies if it cannot write to a log file, just
3460 as it does if it cannot open a log file.
3462 39. Modified OS/Makefile-Linux so that it now contains
3464 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3466 The two -D definitions ensure that Exim is compiled with large file
3467 support, which makes it possible to handle log files that are bigger than
3470 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3471 instance) a domain was checked against a named list that involved a lookup,
3472 causing $domain_data to be set, then another domain was checked against the
3473 same list, then the first domain was re-checked, the value of $domain_data
3474 after the final check could be wrong. In particular, if the second check
3475 failed, it could be set empty. This bug probably also applied to
3478 41. The strip_trailing_dot option was not being applied to the address given
3479 with the -f command-line option.
3481 42. The code for reading a message's header from the spool was incrementing
3482 $received_count, but never initializing it. This meant that the value was
3483 incorrect (doubled) while delivering a message in the same process in which
3484 it was received. In the most common configuration of Exim, this never
3485 happens - a fresh exec is done - but it can happen when
3486 deliver_drop_privilege is set.
3488 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3489 it now includes up to 150 characters of the unexpected data in the log
3492 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3493 and building data strings. The size of both of these buffers was 10 000
3494 bytes - far larger than anybody would *ever* want, thought I. Needless to
3495 say, somebody hit the limit. I have increased the maximum line length to
3496 20 000 and the maximum data length of concatenated lines to 100 000. I have
3497 also fixed two bugs, because there was no checking on these buffers. Tsk,
3498 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3499 buffer is too small.
3501 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3502 lsearch lookups. Now it does.
3504 46. When parsing a route_list item in a manualroute router, a fixed-length
3505 buffer was used for the list of hosts. I made this 1024 bytes long,
3506 thinking that nobody would ever have a list of hosts that long. Wrong.
3507 Somebody had a whole pile of complicated expansion conditions, and the
3508 string was silently truncated, leading to an expansion error. It turns out
3509 that it is easier to change to an unlimited length (owing to other changes
3510 that have happened since this code was originally written) than to build
3511 structure for giving a limitation error. The length of the item that
3512 expands into the list of hosts is now unlimited.
3514 47. The lsearch lookup could not handle data where the length of text line was
3515 more than 4095 characters. Such lines were truncated, leading to shortened
3516 data being returned. It should now handle lines of any length.
3518 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3519 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3522 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3523 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3524 no longer happens in comment lines. A list of replacements is now placed
3525 at the head of all of the source files, except those whose only change is
3526 to replace PERL_COMMAND in the very first #! line.
3528 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3529 messages on the queue, with a bottom-up merge sort, using code contributed
3530 by Michael Haardt. This should make operations like -bp somewhat faster on
3531 large queues. It won't affect queue runners, except when queue_run_in_order
3534 51. Installed eximstats 1.31 in the distribution.
3536 52. Added support for SRV lookups to the dnslookup router.
3538 53. If an ACL referred to $message_body or $message_body_end, the value was not
3539 reset for any messages that followed in the same SMTP session.
3541 54. The store-handling optimization for building very long strings was not
3542 differentiating between the different store pools. I don't think this
3543 actually made any difference in practice, but I've tidied it.
3545 55. While running the routers to verify a sender address, $sender_address
3546 was still set to the sender address. This is wrong, because when routing to
3547 send a bounce to the sender, it would be empty. Therefore, I have changed
3548 it so that, while verifying a sender address, $sender_address is set to <>.
3549 (There is no change to what happens when verifying a recipient address.)
3551 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3552 target A or AAAA records (if not already returned) without resetting the
3553 qualify_single or search_parents options of the DNS resolver. These are
3554 inappropriate in this case because the targets of MX and SRV records must
3555 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3556 target that, when qualified, matched something in the local domain. These
3557 two options are now turned off when doing these lookups.
3559 57. It seems that at least some releases of Reiserfs (which does not have the
3560 concept of a fixed number of inodes) returns zero and not -1 for the
3561 number of available inodes. This interacted badly with check_spool_inodes,
3562 which assumed that -1 was the "no such thing" setting. What I have done is
3563 to check that the total number of inodes is greater than zero before doing
3564 the test of how many are available.
3566 58. When a "warn" ACL statement has a log_message modifier, the message is
3567 remembered, and not repeated. This is to avoid a lot of repetition when a
3568 message has many recipients that cause the same warning to be written.
3569 Howewer, Exim was preserving the list of already written lines for an
3570 entire SMTP session, which doesn't seem right. The memory is now reset if a
3571 new message is started.
3573 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3574 debugging output unless log_rewrite was also set.
3576 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3577 of (int)(handle) when we know that handle contains (void *)(-1).
3579 61. The Exim daemon panic-logs an error return when it closes the incoming
3580 connection. However "connection reset by peer" seems to be common, and
3581 isn't really an error worthy of noting specially, so that particular error
3584 62. When Exim is trying to find all the local interfaces, it used to panic and
3585 die if the ioctl to get the interface flags failed. However, it seems that
3586 on at least one OS (Solaris 9) it is possible to have an interface that is
3587 included in the list of interfaces, but for which you get a failure error
3588 for this call. This happens when the interface is not "plumbed" into a
3589 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3590 failure of the "get flags" call assumes that the interface is down.
3592 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3593 makes life easier for people who are doing arithmetic on fields extracted
3594 from dates, where you often get leading zeros that should not be
3595 interpreted as octal.
3597 64. Added qualify_domain to the redirect router, to override the global
3600 65. If a pathologically long header line contained very many addresses (the
3601 report of this problem mentioned 10 000) and each of them was rewritten,
3602 Exim could use up a very large amount of memory. (It kept on making new
3603 copies of the header line as it rewrote, and never released the old ones.)
3604 At the expense of a bit more processing, the header rewriting function has
3605 been changed so that it no longer eats memory in this way.
3607 66. The generation of the Received: header has been moved from the time that a
3608 message starts to be received, to the time that it finishes. The timestamp
3609 in the Received: header should now be very close to that of the <= log
3610 line. There are two side-effects of this change:
3612 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3613 logged header lines no longer include the local Received: line, because
3614 it has not yet been created. The same applies to a copy of the message
3615 that is returned to a non-SMTP sender when a message is rejected.
3617 (b) When a filter file is tested using -bf, no additional Received: header
3618 is added to the test message. After some thought, I decided that this
3621 This change does not affect the value of $received_for. It is still set
3622 after address rewriting, but before local_scan() is called.
3624 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3626 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3627 gave an unhelpful panic error message, and a defer error. I have managed to
3628 change this behaviour so that it now rejects any supplied certificate,
3629 which seems right, as the list of acceptable certificates is empty.
3631 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3632 gave an unhelpful defer error. I have not managed to make this reject any
3633 supplied certificates, but the error message it gives is "no certificate
3634 supplied", which is not helpful.
3636 70. exigrep's output now also includes lines that are not associated with any
3637 message, but which match the given pattern. Implemented by a patch from
3638 Martin Sluka, which also tidied up the Perl a bit.
3640 71. Recipient callout verification, like sender verification, was using <> in
3641 the MAIL FROM command. This isn't really the right thing, since the actual
3642 sender may affect whether the remote host accepts the recipient or not. I
3643 have changed it to use the actual sender in the callout; this means that
3644 the cache record is now keyed on a recipient/sender pair, not just the
3645 recipient address. There doesn't seem to be a real danger of callout loops,
3646 since a callout by the remote host to check the sender would use <>.
3647 [SEE ABOVE: changed after hitting problems.]
3649 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3650 temporary errors. However, in the case of such a code being given after
3651 the end of a data transmission (i.e. after ".") Exim was failing to write
3652 a retry record for the message. (Yes, there was some broken host that was
3653 actually sending 8xx at this point.)
3655 73. An unknown lookup type in a host list could cause Exim to panic-die when
3656 the list was checked. (An example that provoked this was putting <; in the
3657 middle of a list instead of at the start.) If this happened during a DATA
3658 ACL check, a -D file could be left lying around. This kind of configuration
3659 error no longer causes Exim to die; instead it causes a defer errror. The
3660 incident is still logged to the main and panic logs.
3662 74. Buglet left over from Exim 3 conversion. The message "too many messages
3663 in one connection" was written to the rejectlog but not the mainlog, except
3664 when address rewriting (yes!) was being logged.
3666 75. Added write_rejectlog option.
3668 76. When a system filter was run not as root (that is, when system_filter_user
3669 was set), the values of the $n variables were not being returned to the
3670 main process; thus, they were not subsequently available in the $sn
3673 77. Added +return_path_on_delivery log selector.
3675 78. A connection timeout was being treated differently from recipients deferred
3676 when testing hosts_max_try with a message that was older than the host's
3677 retry timeout. (The host should not be counted, thus allowing all hosts to
3678 be tried at least once before bouncing.) This may have been the cause of an
3679 occasionally reported bug whereby a message would remain on the queue
3680 longer than the retry timeout, but would be bounced if a delivery was
3681 forced. I say "may" because I never totally pinned down the problem;
3682 setting up timeout/retry tests is difficult. See also the next item.
3684 79. The ultimate address timeout was not being applied to errors that involved
3685 a combination of host plus message (for example, a timeout on a MAIL
3686 command). When an address resolved to a number of possible hosts, and they
3687 were not all tried for each delivery (e.g. because of hosts_max_try), a
3688 message could remain on the queue longer than the retry timeout.
3690 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3693 81. Fixed an obscure SMTP outgoing bug which required at least the following
3694 conditions: (a) there was another message waiting for the same server;
3695 (b) the server returned 5xx to all RCPT commands in the first message so
3696 that the message was not completed; (c) the server dropped the connection
3697 or gave a negative response to the RSET that Exim sends to abort the
3698 transaction. The observed case was a dropped connection after DATA that had
3699 been sent in pipelining mode. That is, the server had advertised PIPELINING
3700 but was not implementing it correctly. The effect of the bug was incorrect
3701 behaviour, such as trying another host, and this could lead to a crash.
3707 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3708 and daemon.c were passed as pointers to ints; they should have been
3709 pointers to socklen_t variables (which are typically unsigned ints).
3711 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3714 3. Fixed a really odd bug that affected only the testing scheme; patching a
3715 certain fixed string in the binary changed the value of another string that
3716 happened to be identical to the end of the original first string.
3718 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3719 name", it returns that address as the IP address. On some operating
3720 systems (e.g. Solaris), it also passes back the IP address string as the
3721 "host name". However, on others (e.g. Linux), it passes back an empty
3722 string. Exim wasn't checking for this, and was changing the host name to an
3723 empty string, assuming it had been canonicized.
3725 5. Although rare, it is permitted to have more than one PTR record for a given
3726 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3727 all the names associated with an address, because they do in Solaris.
3728 However, it seems that they do not in Linux for data that comes from the
3729 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3730 I found this out when I moved to a new Linux workstation and tried to run
3731 the Exim test suite.
3733 To get round this problem I have changed the code so that it now does its
3734 own call to the DNS to look up PTR records when searching for a host name.
3735 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3736 addresses that are only in /etc/hosts are still found.
3738 This behaviour is, however, controlled by an option called host_lookup_
3739 order, which defaults to "bydns:byaddr". If people want to use the other
3740 order, or indeed, just use one or the other means of lookup, they can
3741 specify it in this variable.
3743 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3744 some operating systems, this comes back from gethostbyaddr() as an empty
3745 string, and this is what Exim used to test for. However, it seems that in
3746 other systems, "." is yielded. Exim now tests for this case too.
3748 7. The values of check_spool_space and check_log_space are now held internally
3749 as a number of kilobytes instead of an absolute number of bytes. If a
3750 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3751 kilobyte. This means that much larger values can be stored.
3753 8. Exim monitor: an attempt to get the action menu when not actually pointing
3754 at a message produces an empty menu entitled "No message selected". This
3755 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3756 no entries in it ("Shell widget menu has zero width and/or height"). So I
3757 have added a single, blank menu entry in this case.
3759 9. Added ${quote_local_part.
3761 10. MIME decoding is now applied to the contents of Subject: header lines when
3764 11. Now that a reference to $sender_host_address automatically causes a reverse
3765 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3766 host lookup before query-style lookups in lists that might use this
3767 variable. This has therefore been abolished, and the "net-" prefix is no
3768 longer necessary for query-style lookups.
3770 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3771 have been a typo for LFLAGS, so it has been changed.
3773 13. The install script calls Exim with "-C /dev/null" in order to find the
3774 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3775 to be output. Howeve, since Exim outputs its version number before the
3776 error, it didn't break the script. It just looked ugly. I fixed this by
3777 always allowing "-C /dev/null" if the caller is root.
3779 14. Ignore overlarge ACL variable number when reading spool file - insurance
3780 against a later release with more variables having written the file.
3782 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3783 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3784 this, as well as the form without the "IPv6" on the front.
3786 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3787 OS/Makefile-Darwin file.
3789 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3790 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3792 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3793 variable. However, some OS use ints for inodes. I've added cast to long int
3794 to get rid of the compiler warning.
3796 19. I had forgotten to lock out "/../" in configuration file names when
3797 ALT_CONFIG_PREFIX was set.
3799 20. Routers used for verification do not need to specify transports. However,
3800 if such a router generated a host list, and callout was configured, Exim
3801 crashed, because it could not find a port number from the (non-existent)
3802 transport. It now assumes port 25 in this circumstance.
3804 21. Added the -t option to exigrep.
3806 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3807 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3808 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3809 them is set, LOOKUP_LSEARCH is forced.
3811 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3813 24. Added sender and host information to the "rejected by local_scan()" log
3814 line; previously there was no indication of these.
3816 25. Added .include_if_exists.
3818 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3819 Linux. It turns out that not all file systems support this. Apparently some
3820 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3821 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3822 not supported on the file descriptor, is now ignored when Exim is trying to
3823 sync a directory. This applies only to Linux.
3825 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3827 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3828 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3831 29. Added support for maildirsize files from supplied patch (modified a bit).
3833 30. The use of :fail: followed by an empty string could lead Exim to respond to
3834 sender verification failures with (e.g.):
3836 550 Verification failed for <xxx>
3837 550 Sender verify failed
3839 where the first response line was missing the '-' that indicates it is not
3840 the final line of the response.
3842 31. The loop for finding the name of the user that called Exim had a hardwired
3843 limit of 10; it now uses the value of finduser_retries, which is used for
3844 all other user lookups.
3846 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3849 33. Exim was neglecting to zero errno before one call of strtol() when
3850 expanding a string and expecting an integer value. On some systems this
3851 resulted in spurious "integer overflow" errors. Also, it was casting the
3852 result into an int without checking.
3854 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3855 did not work. The code looks as if it has *never* worked, though it appears
3856 to have been documented since at least releast 1.62. I have made it work.
3858 35. The "timeout_DNS" error in retry rules, also documented since at least
3859 1.62, also never worked. As it isn't clear exactly what this means, and
3860 clearly it isn't a major issue, I have abolished the feature by treating it
3861 as "timeout", and writing a warning to the main and panic logs.
3863 36. The display of retry rules for -brt wasn't always showing the error code
3866 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3867 timeout_connect_A, timeout_connect_MX.
3869 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3870 to the empty sender.
3872 39. The daemon was not analysing the content of -oX till after it had closed
3873 stderr and disconnected from the controlling terminal. This meant that any
3874 syntax errors were only noted on the panic log, and the return code from
3875 the command was 0. By re-arranging the code a little, I've made the
3876 decoding happen first, so such errors now appear on stderr, and the return
3877 code is 1. However, the actual setting up of the sockets still happens in
3878 the disconnected process, so errors there are still only recorded on the
3881 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3882 connections (as happens on some IP stacks) was logged at start up time as
3883 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3884 it from "IPv6 and IPv4", which means that two separate sockets are being
3887 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3888 says whether AF_INET or AF_INET6 was passed as an argument.
3890 42. Exiwhat output was messed up when time zones were included in log
3893 43. Exiwhat now gives more information about the daemon's listening ports,
3894 and whether -tls-on-connect was used.
3896 44. The "port" option of the smtp transport is now expanded.
3898 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3899 silently ignored. Now an error message is written to the main and panic
3902 46. There's a new ACL modifier called "logwrite" which writes to a log file
3903 as soon as it is encountered.
3905 47. Added $local_user_uid and $local_user_gid at routing time.
3907 48. Exim crashed when trying to verify a sender address that was being
3910 49. Exim was recognizing only a space character after ".include". It now also
3911 recognizes a tab character.
3913 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3914 extracting the relevant information from the specification. The man page no
3915 longer contains scrambled data for the -d option, and I've added a section
3916 at the front about calling Exim under different names.
3918 51. Added "extra_headers" argument to the "mail" command in filter files.
3920 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3923 53. Installed eximstats 1.29.
3925 54. Added transport_filter_timeout as a generic transport option.
3927 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3928 Cc: header lines. This was required by RFC 822, but it not required by RFC
3931 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3932 incoming messages that did not have them. Now it does so only if the
3933 message originates locally, that is, if there is no associated remote host
3934 address. When Resent- header lines are present, this applies to the Resent-
3935 lines rather than the non-Resent- lines.
3937 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3938 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3940 58. Messages for configuration errors now include the name of the main
3941 configuration file - useful now that there may be more than one file in a
3942 list (.included file names were always shown).
3944 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3945 for those rare installations that do not start the daemon as root or run it
3946 setuid root. I've cut out the call to initgroups() if the daemon is not
3949 60. The Exim user and group can now be bound into the binary as text strings
3950 that are looked up at the start of Exim's processing.
3952 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3954 62. Added $mailstore_basename variable.
3956 63. Installed patch to sieve.c from Michael Haardt.
3958 64. When Exim failed to open the panic log after failing to open the main log,
3959 the original message it was trying to log was written to stderr and debug
3960 output, but if they were not available (the usual case in production), it
3961 was lost. Now it is written to syslog before the two lines that record the
3962 failures to open the logs.
3964 65. Users' Exim filters run in subprocesses under the user's uid. It is
3965 possible for a "deliver" command or an alias in a "personal" command to
3966 provoke an address rewrite. If logging of address rewriting is configured,
3967 this fails because the process is not running as root or exim. There may be
3968 a better way of dealing with this, but for the moment (because 4.30 needs
3969 to be released), I have disabled address rewrite logging when running a
3970 filter in a non-root, non-exim process.
3976 1. The buildconfig auxiliary program wasn't quoting the value set for
3977 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3978 not defined. This bug was masked in 4.22 by the effect that was fixed in
3981 2. Some messages that were rejected after a message id was allocated were
3982 shown as "incomplete" by exigrep. It no longer does this for messages that
3983 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3985 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3986 have allow_domain_literals set, the ID did not get logged in the <= line.
3987 Domain literals are now always recognized in Message-ID: header lines.
3989 4. The first argument for a ${extract expansion item is the key name or field
3990 number. Leading and trailing spaces in this item were not being ignored,
3991 causing some misleading effects.
3993 5. When deliver_drop_privilege was set, single queue runner processes started
3994 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3995 same command in the process it spins off) were not dropping privilege.
3997 6. When the daemon running as "exim" started a queue runner, it always
3998 re-executed Exim in the spun-off process. This is a waste of effort when
3999 deliver_drop_privilege is set. The new process now just calls the
4000 queue-runner function directly.
4006 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4009 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4010 ignored. Though the use of -oP was forcing the writing of a pid file, it
4011 was always written to the default place.
4013 3. If the message "no IP address found for host xxxx" is generated during
4014 incoming verification, it is now followed by identification of the incoming
4015 connection (so you can more easily find what provoked it).
4017 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4019 5. Added some features to "harden" Exim a bit more against certain attacks:
4021 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4022 be put in Local/Makefile. This is like the never_users runtime option,
4023 but it cannot be overridden. The default setting is "root".
4025 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4026 prefix string with which any file named in a -C command line option
4029 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4030 is retained for -C and -D only if the caller of Exim is root. Without
4031 it, the exim user may also use -C and -D and retain privilege.
4033 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4034 command line option is disabled.
4036 6. Macro names set by the -D option must start with an upper case letter, just
4037 like macro names defined in the configuration file.
4039 7. Added "dereference=" facility to LDAP.
4041 8. Two instances of the typo "uknown" in the source files are fixed.
4043 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4044 the Configure-Makefile script screwed up while processing it.
4046 10. Incorporated PCRE 4.4.
4048 11. The SMTP synchronization check was not operating right at the start of an
4049 SMTP session. For example, it could not catch a HELO sent before the client
4050 waited for the greeting. There is now a check for outstanding input at the
4051 point when the greeting is written. Because of the duplex, asynchronous
4052 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4053 way, but not yet received, when the check is performed.
4055 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4056 on TCP/IP sockets, because this apparently causes some broken clients to
4059 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4060 unchanged) from the Cygwin maintainer.
4062 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4063 was supported instead of testing for mbx. Effectively a typo.
4065 15. The spa authenticator server code was not checking that the input it
4066 received was valid base64.
4068 16. The debug output line for the "set" modifier in ACLs was not showing the
4069 name of the variable that was being set.
4071 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4073 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4074 lookup on its own. Something else was needed to trigger the lookup. For
4075 example, a match in host_lookup or the need for a host name in a host list.
4076 Now, if $sender_host_name is referenced and the host name has not yet been
4077 looked up, a lookup is performed. If the lookup fails, the variable remains
4078 empty, and $host_lookup_failed is set to "1".
4080 19. Added "eqi" as a case-independent comparison operator.
4082 20. The saslauthd authentication condition could segfault if neither service
4083 nor realm was specified.
4085 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4086 error message that was logged was misleading, and incoming SMTP
4087 connections were dropped. The message is now more accurate, and temporary
4088 errors are given to SMTP connections.
4090 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4091 (except RSET) with a 421 error, until QUIT is received. However, it was
4092 failing to send a response to QUIT.
4094 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4095 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4096 ACL is now run after the helo_try_verify_hosts code.
4098 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4099 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4100 case-independent, so other case variants are also recognized). Apparently
4101 some systems use these upper case variants.
4103 25. If more than two messages were waiting for the same host, and a transport
4104 filter was specified for the transport, Exim sent two messages over the
4105 same TCP/IP connection, and then failed with "socket operation on non-
4106 socket" when it tried to send the third.
4108 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4110 27. The extern definition of crypt16() in expand.c was not being excluded when
4111 the OS had its own crypt16() function.
4113 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4114 as a preferred synonym for return_size_limit, both as an option and as an
4117 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4119 30. Changed the default configuration ACL to relax the local part checking rule
4120 for addresses that are not in any local domains. For these addresses,
4121 slashes and pipe symbols are allowed within local parts, but the sequence
4122 /../ is explicitly forbidden.
4124 31. SPA server authentication was not clearing the challenge buffer before
4127 32. log_message in a "warn" ACL statement was writing to the reject log as
4128 well as to the main log, which contradicts the documentation and doesn't
4129 seem right (because no rejection is happening). So I have stopped it.
4131 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4132 However, I am unable to do any testing of this.
4134 34. Fixed an infelicity in the appendfile transport. When checking directories
4135 for a mailbox, to see if any needed to be created, it was accidentally
4136 using path names with one or more superfluous leading slashes; tracing
4137 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4139 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4140 contents are added to the log line that is written for every discarded
4141 recipient. (Previously a log_message setting was ignored.)
4143 36. The ${quote: operator now quotes the string if it is empty.
4145 37. The install script runs exim in order to find its version number. If for
4146 some reason other than non-existence or emptiness, which it checks, it
4147 could not run './exim', it was installing it with an empty version number,
4148 i.e. as "exim-". This error state is now caught, and the installation is
4151 38. An argument was missing from the function that creates an error message
4152 when Exim fails to connect to the socket for saslauthd authentication.
4153 This could cause Exim to crash, or give a corrupted message.
4155 39. Added isip, isip4, and isip6 to ${if conditions.
4157 40. The ACL variables $acl_xx are now saved with the message, and can be
4158 accessed later in routers, transports, and filters.
4160 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4161 strings in the file are not string-expanded.
4163 42. If a MAIL command specified a SIZE value that was too large to fit into an
4164 int variable, the check against message_size_limit failed. Such values are
4165 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4166 day this will have to be increased, but I don't think I want to be around
4167 when emails are that large.
4174 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4175 iconv() is not standard in FreeBSD.
4177 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4178 IPv6 enabled. The observed symptom was a segmentation fault on return from
4179 the function os_common_find_running_interfaces() in src/os.c.
4181 3. In the check_special_case() function in daemon.c I had used "errno" as an
4182 argument name, which causes warnings on some systems. This was basically a
4183 typo, since it was named "eno" in the comments!
4185 4. The code that waits for the clock to tick (at a resolution of some fraction
4186 of a second) so as to ensure message-id uniqueness was always waiting for
4187 at least one whole tick, when it could have waited for less. [This is
4188 almost certainly not relevant at current processor speeds, where it is
4189 unlikely to ever wait at all. But we try to future-proof.]
4191 5. The function that sleeps for a time interval that includes fractions of a
4192 second contained a race. It did not block SIGALRM between setting the
4193 timer, and suspending (a couple of lines later). If the interval was short
4194 and the sigsuspend() was delayed until after it had expired, the suspension
4195 never ended. On busy systems this could lead to processes getting stuck for
4198 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4199 process, before it forks any delivery processes. The open lookup caching
4200 mechanism meant that the open file or database connection was passed into
4201 the delivery process. The problem was that delivery processes always tidy
4202 up cached lookup data. This could cause a problem for the next delivery
4203 process started by the queue runner, because the external queue runner
4204 process does not know about the closure. So the next delivery process
4205 still has data in the lookup cache. In the case of a file lookup, there was
4206 no problem because closing a file descriptor in a subprocess doesn't affect
4207 the parent. However, if the lookup was caching a connection to a database,
4208 the connection was closed, and the second delivery process was likely to
4209 see errors such as "PGSQL: query failed: server closed the connection
4210 unexpectedly". The problem has been fixed by closing all cached lookups
4211 in a queue runner before running a delivery process.
4213 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4214 seem to have the "const" qualifier which it has on other OS. I've
4217 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4218 *of the same type* (client or server) with the same public name that an
4219 error should be diagnosed.
4221 9. When Exim looked up a host name for an IP address, but failed to find the
4222 original IP address when looking up the host name (a safety check), it
4223 output the message "<ip address> does not match any IP for NULL", which was
4224 confusing, to say the least. The bug was that the host name should have
4225 appeared instead of "NULL".
4227 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4228 user that is built into the binary, and the -C or -D options is used, root
4229 privilege is dropped before the configuration file is read. In addition,
4230 logging is switched to stderr instead of the normal log files. If the
4231 configuration then re-defines the Exim user, the unprivileged environment
4232 is probably not what is expected, so Exim logs a panic warning message (but
4235 However, if deliver_drop_privilege is set, the unprivileged state may well
4236 be exactly what is intended, so the warning has been cut out in that case,
4237 and Exim is allowed to try to write to its normal log files.
4243 1. smtp_return_error_details was not giving details for temporary sender
4244 or receiver verification errors.
4246 2. Diagnose a configuration error if two authenticators have the same public
4249 3. Exim used not to create the message log file for a message until the first
4250 delivery attempt. This could be confusing when incoming messages were held
4251 for policy or load reasons. The message log file is now created at the time
4252 the message is received, and an initial "Received" line is written to it.
4254 4. The automatically generated man page for command line options had a minor
4255 bug that caused no ill effects; however, a more serious problem was that
4256 the procedure for building the man page automatically didn't always
4257 operate. Consequently, release 4.20 contains an out-of-date version. This
4258 shouldn't happen again.
4260 5. When building Exim with embedded Perl support, the script that builds the
4261 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4262 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4264 6. The freeze_tell option was not being used for messages that were frozen on
4265 arrival, either by an ACL or by local_scan().
4267 7. Added the smtp_incomplete_transaction log selector.
4269 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4270 was accepting AUTH without a new EHLO.
4272 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4273 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4276 10. Logging of TCP/IP connections (when configured) now happens in the main
4277 daemon process instead of the child process, so that the TCP/IP connection
4278 count is more accurate (but it can never be perfect).
4280 11. The use of "drop" in a nested ACL was not being handled correctly in the
4281 outer ACL. Now, if condition failure induced by the nested "drop" causes
4282 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4283 or "require"), the connection is dropped.
4285 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4286 that yield "discard" can now be used with an "accept" or a "discard" verb,
4287 but an error is generated for any others (because I can't see a useful way
4288 to define what should happen).
4290 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4291 are now processed in the same way as lines in the Exim configuration file.
4292 In particular, continuation lines are supported.
4294 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4296 15. Added -ti meaning -t -i.
4298 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4299 domains, and warn by logging if others are found.
4301 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4302 returned from GIFCONF when Exim is trying to find the list of interfaces on
4303 a host. The code in os.c has been modified to copy each ifreq to an aligned
4304 structure in all cases.
4306 Also, in some cases, the returned ifreq's were being copied to a 'struct
4307 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4308 means the last couple of bytes of an IPv6 address could be chopped if the
4309 ifreq contained only a normal sockaddr (14 bytes storage).
4311 18. Named domain lists were not supported in the hosts_treat_as_local option.
4312 An entry such as +xxxx was not recognized, and was treated as a literal
4315 19. Ensure that header lines added by a DATA ACL are included in the reject log
4316 if the ACL subsequently rejects the message.
4318 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4319 MD5 (which is deprecated).
4321 21. When testing a filter file using -bf, Exim was writing a message when it
4322 took the sender from a "From " line in the message, but it was not doing so
4323 when it took $return_path from a Return-Path: header line. It now does.
4325 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4326 with a valid header line field name (a series of printing characters
4327 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4329 23. Changed "disc" in the source to "disk" to conform to the documentation and
4330 the book and for uniformity.
4332 24. Ignore Sendmail's -Ooption=value command line item.
4334 25. When execve() failed while trying to run a command in a pipe transport,
4335 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4336 could be confused with a return value of 69 from the command itself. This
4337 has been changed to 127, the value the shell returns if it is asked to run
4338 a non-existent command. The wording for the related log line suggests a
4339 non-existent command as the problem.
4341 26. If received_header_text expands to an empty string, do not add a Received:
4342 header line to the message. (Well, it adds a token one on the spool, but
4343 marks it "old" so that it doesn't get used or transmitted.)
4345 27. Installed eximstats 1.28 (addition of -nt option).
4347 28. There was no check for failure on the call to getsockname() in the daemon
4348 code. This can fail if there is a shortage of resources on the system, with
4349 ENOMEM, for example. A temporary error is now given on failure.
4351 29. Contrary to the C standard, it seems that in some environments, the
4352 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4353 program. Exim now does this explicitly; it affects the formatting of
4354 timestamps using strftime().
4356 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4357 complaints. I've now initialized all the variables, to avoid this.
4359 32. Header lines added by a system filter were not being "seen" during
4360 transport-time rewrites.
4362 33. The info_callback() function passed to OpenSSL is set up with type void
4363 (*)(SSL *, int, int), as described somewhere. However, when calling the
4364 function (actually a macro) that sets it up, the type void(*)() is
4365 expected. I've put in a cast to prevent warnings from picky compilers.
4367 34. If a DNS black list lookup found a CNAME record, but there were no A
4368 records associated with the domain it pointed at, Exim crashed.
4370 35. If a DNS black list lookup returned more than one A record, Exim ignored
4371 all but the first. It now scans all returned addresses if a particular IP
4372 value is being sought. In this situation, the contents of the
4373 $dnslist_value variable are a list of all the addresses, separated by a
4376 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4377 to accept a host name and all its aliases if the forward lookup for any of
4378 them yielded the IP address of the incoming connection. Now it accepts only
4379 those names whose forward lookup yields the correct IP address. Any other
4380 names are discarded. This closes a loophole whereby a rogue DNS
4381 administrator could create reverse DNS records to break through a
4382 wildcarded host restriction in an ACL.
4384 37. If a user filter or a system filter that ran in a subprocess used any of
4385 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4386 the wrong values were passed to the pipe command ($thisaddress had the
4387 value of $0, $0 had the value of $1, etc). This bug was introduced by
4388 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4390 38. Improved the line breaking for long SMTP error messages from ACLs.
4391 Previously, if there was no break point between 40 and 75 characters, Exim
4392 left the rest of the message alone. Two changes have been made: (a) I've
4393 reduced the minimum length to 35 characters; (b) if it can't find a break
4394 point between 35 and 75 characters, it looks ahead and uses the first one
4395 that it finds. This may give the occasional overlong line, but at least the
4396 remaining text gets split now.
4398 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4399 file descriptors might be low, and that setting 1000 would always raise it.
4400 It turns out that in some environments, the limit is already over 1000 and
4401 that lowering it causes trouble. So now Exim takes care not to decrease it.
4403 40. When delivering a message, the value of $return_path is set to $sender_
4404 address at the start of routing (routers may change the value). By an
4405 oversight, this default was not being set up when an address was tested by
4406 -bt or -bv, which affected the outcome if any router or filter referred to
4409 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4410 log only when "message" or "log_message" are set. However, if one of the
4411 conditions was an address verification, or a call to a nested ACL, the
4412 messages generated by the underlying test were being passed through. This
4413 no longer happens. The underlying message is available in $acl_verify_
4414 message for both "message" and "log_message" expansions, so it can be
4415 passed through if needed.
4417 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4418 new expansion $bh_ to give the encoded byte string without charset
4419 translation. Translation happens only if iconv() is available; HAVE_ICONV
4420 indicates this at build time. HEADERS_CHARSET gives the charset to
4421 translate to; headers_charset can change it in the configuration, and
4422 "headers charset" can change it in an individual filter file.
4424 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4425 that creates RFC 2047 encoded "words" labels them as that charset instead
4426 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4427 expansion operator; (ii) when Exim creates a From: line for a local
4428 message; (iii) when a header line is rewritten to include a "phrase" part.
4430 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4431 buggy, causing it to skip the first lines of messages whose message ID
4432 ended in 'D'. This would not have bitten before Exim release 4.14, because
4433 message IDs were unlikely to end in 'D' before then. The effect was to have
4434 incorrect size information for certain domains.
4436 45. #include "config.h" was missing at the start of the crypt16.c module. This
4437 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4440 46. If there was a timeout during a "random" callout check, Exim treated it as
4441 a failure of the random address, and carried on sending RSET and the real
4442 address. If the delay was just some slowness somewhere, the response to the
4443 original RCPT would be taken as a response to RSET and so on, causing
4444 mayhem of various kinds.
4446 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4447 when I implemented it. It didn't allow for the fact that some option values
4448 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4449 the right test for positive values.
4451 48. Domain names in DNS records are case-independent. Exim always looks them up
4452 in lower case. Some resolvers return domain names in exactly the case they
4453 appear in the zone file, that is, they may contain uppercase letters. Not
4454 all resolvers do this - some return always lower case. Exim was treating a
4455 change of case by a resolver as a change of domain, similar to a widening
4456 of a domain abbreviation. This triggered its re-routing code and so it was
4457 trying to route what was effectively the same domain again. This normally
4458 caused routing to fail (because the router wouldn't handle the domain
4459 twice). Now Exim checks for this case specially, and just changes the
4460 casing of the domain that it ultimately uses when it transmits the message
4463 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4466 50. If a filter generated a file delivery with a non-absolute name (possible if
4467 no home directory exists for the router), the forbid_file option was not
4470 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4471 the existing equality matching.
4473 52. Exim was using ints instead of ino_t variables in some places where it was
4474 dealing with inode numbers.
4476 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4477 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4478 called TMPDIR, and if it finds it is different, it changes its value.
4480 54. The smtp_printf() function is now made available to local_scan() so
4481 additional output lines can be written before returning. There is also an
4482 smtp_fflush() function to enable the detection of a dropped connection.
4483 The variables smtp_input and smtp_batched_input are exported to
4486 55. Changed the default runtime configuration: the message "Unknown user"
4487 has been removed from the ACL, and instead placed on the localuser router,
4488 using the cannot_route_message feature. This means that any verification
4489 failures that generate their own messages won't get overridden. Similarly,
4490 the "Unrouteable address" message that was in the ACL for unverifiable
4491 relay addresses has also been removed.
4493 56. Added hosts_avoid_esmtp to the smtp transport.
4495 57. The exicyclog script was not checking for the esoteric option
4496 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4497 will work only if exicyclog is run under the appropriate euid.
4499 58. Following a discussion on the list, the rules by which Exim recognises line
4500 endings on incoming messages have been changed. The -dropcr and drop_cr
4501 options are now no-ops, retained only for backwards compatibility. The
4502 following line terminators are recognized: LF CRLF CR. However, special
4503 processing applies to CR:
4505 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4506 nor a local message in the state where . is a terminator.
4508 (ii) If a bare CR is encountered in a header line, an extra space is added
4509 after the line terminator so as not to end the header. The reasoning
4510 behind this is that bare CRs in header lines are most likely either
4511 to be mistakes, or people trying to play silly games.
4513 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4514 was being incorrectly given as 18 bytes larger than it should have been.
4515 This is a VOB (very old bug).
4517 60. This may never have affected anything current, but just in case it has:
4518 When the local host is found other than at the start of a list of hosts,
4519 the local host, those with the same MX, and any that follow, are discarded.
4520 When the list in question was part of a longer list of hosts, the following
4521 hosts (not currently being processed) were also being discarded. This no
4522 longer happens. I'm not sure if this situation could ever has previously
4525 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4528 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4529 header. This is something that is recommended in a new Internet Draft, and
4530 is something that is documented as being done by Sendmail. There are two
4531 possible values. For messages generated by the autoreply transport, Exim
4534 Auto-Submitted: auto-replied
4536 whereas for all other generated messages (e.g. bounces) it adds
4538 Auto-Submitted: auto-generated
4540 63. The "personal" condition in filters now includes a test for the
4541 Auto-Submitted: header. If it contains the string "auto-" the message it
4542 not considered personal.
4544 64. Added rcpt_include_affixes as a generic transport option.
4546 65. Added queue_only_override (default true).
4548 66. Added the syslog_duplication option.
4550 67. If what should have been the first header line of a message consisted of
4551 a space followed by a colon, Exim was mis-interpreting it as a header line.
4552 It isn't of course - it is syntactically invalid and should therefore be
4553 treated as the start of the message body. The misbehaviour could have
4554 caused a number of strange effects, including loss of data in subsequent
4555 header lines, and spool format errors.
4557 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4558 client host had authenticated. This control can now be exercised by an ACL
4559 for more flexibility.
4561 69. By default, callouts do not happen when testing with -bh. There is now a
4562 variant, -bhc, which does actually run the callout code, including
4563 consulting and updating the callout cache.
4565 70. Added support for saslauthd authentication, courtesy of Alexander
4568 71. If statvfs() failed on the spool or log directories while checking their
4569 size for availability, Exim confusingly gave the error "space shortage".
4570 Furthermore, in debugging mode it crashed with a floating point exception.
4571 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4572 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4573 serious problem, Exim now writes to the main and panic logs when this
4574 happens, with details of the failure. It then refuses to accept the
4575 incoming message, giving the message "spool directory problem" or "log
4576 directory problem" with a 421 code for SMTP messages.
4578 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4579 0, 1, and 2 exist, because some OS complain for execs without them (see
4580 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4581 descriptors. However, the code omitted to check that the open succeeded,
4582 causing mysterious errors if for some reason the permissions on /dev/null
4583 got screwed. Now Exim writes a message to the main and panic logs, and
4584 bombs out if it can't open /dev/null.
4586 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4587 interact so that it is all more flexible. It is supposed to remain
4588 backwards compatible. Also added extra_local_interfaces.
4590 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4591 to bomb out with an assertion failure - to the client this appears as a
4592 connection drop. This problem occurs in the part of the code that was taken
4593 from the Samba project. Fortunately, the assertion is in a very simple
4594 function, so I have fixed this by reproducing the function inline in the
4595 one place where it is called, and arranging for authentication to fail
4596 instead of killing the process with assert().
4598 75. The SPA client code was not working when the server requested OEM rather
4599 than Unicode encoding.
4601 76. Added code to make require_files with a specific uid setting more usable in
4602 the case where statting the file as root fails - usually a non-root-mounted
4603 NFS file system. When this happens and the failure is EACCES, Exim now
4604 forks a subprocess and does the per-uid checking as the relevant uid.
4606 77. Added process_log_path.
4608 78. If log_file_path was not explicitly set, a setting of check_log_space or
4609 check_log_inodes was ignored.
4611 79. If a space check for the spool or log partitions fails, the incident is now
4612 logged. Of course, in the latter case the data may get lost...
4614 80. Added the %p formatting code to string_format() so that it can be used to
4615 print addresses in debug_print(). Adjusted all the address printing in the
4616 debugging in store.c to use %p rather than %d.
4618 81. There was a concern that a line of code in smtp_in.c could overflow a
4619 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4620 initially expressed, the concern was not well-founded, because trailing
4621 spaces are removed early. However, if the trailing spaces were followed by
4622 a NULL, they did not get removed, so the overflow was possible. Two fixes
4625 (a) I re-wrote the offending code in a cleaner fashion.
4626 (b) If an incoming SMTP command contains a NULL character, it is rejected
4629 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4630 runs initgroups(), so that if the Exim user is in any additional groups,
4631 they will be used during message reception.
4637 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob