1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.476 2007/02/07 14:41:13 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".
104 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
105 fixed by 4.65/MH/01 (is this a record?) are fixed:
107 (i) An empty string was always treated as zero by the numeric comparison
108 operators. This behaviour has been restored.
110 (ii) It is documented that the numeric comparison operators always treat
111 their arguments as decimal numbers. This was broken in that numbers
112 starting with 0 were being interpreted as octal.
114 While fixing these problems I realized that there was another issue that
115 hadn't been noticed. Values of message_size_limit (both the global option
116 and the transport option) were treated as octal if they started with 0.
117 The documentation was vague. These values are now always treated as
118 decimal, and I will make that clear in the documentation.
124 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
125 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
128 MH/01 Don't check that the operands of numeric comparison operators are
129 integers when their expansion is in "skipping" mode (fixes bug
130 introduced by 4.64-PH/07).
132 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
133 child addresses, Exim now panics and dies. Previously, because the count
134 is held in a short int, deliveries were likely to be lost. As such a
135 large number of recipients for a single message is ridiculous
136 (performance will be very, very poor), I have chosen to impose a limit
137 rather than extend the field.
143 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
144 leftover -K file (the existence of which was triggered by #402).
145 While we were at it, introduced process PID as part of the -K
146 filename. This should rule out race conditions when creating
149 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
150 processing considerably. Previous code took too long for large mails,
151 triggering a timeout which in turn triggers #401.
153 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
154 in the DK code in transports.c. sendfile() is not really portable,
155 hence the _LINUX specificness.
157 TF/01 In the add_headers option to the mail command in an Exim filter,
158 there was a bug that Exim would claim a syntax error in any
159 header after the first one which had an odd number of characters
162 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
163 callout verification, Exim cached a "reject" for the entire domain. This
164 is correct for most verifications, but it is not correct for a recipient
165 verification with use_sender or use_postmaster set, because in that case
166 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
167 case of MAIL FROM:<> rejection from other early rejections (e.g.
168 rejection of HELO). When verifying a recipient using a non-null MAIL
169 address, the cache is ignored if it shows MAIL FROM:<> rejection.
170 Whatever the result of the callout, the value of the domain cache is
171 left unchanged (for any other kind of callout, getting as far as trying
172 RCPT means that the domain itself is ok).
174 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
177 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
178 manifest itself as EPIPE rather than ECONNECT. When tidying away a
179 session, the daemon ignores ECONNECT errors and logs others; it now
180 ignores EPIPE as well.
182 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
183 (quoted-printable decoding).
185 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
186 later the small subsequent patch to fix an introduced bug.
188 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
190 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
192 PH/08 An error is now given if message_size_limit is specified negative.
194 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
195 to be given (somewhat) arbitrary names.
197 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
200 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
201 miscellaneous code fixes
203 PH/10 Added the log_reject_target ACL modifier to specify where to log
206 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
207 hostname. This is wrong, because it relates to the incoming message (and
208 probably the interface on which it is arriving) and not to the outgoing
209 callout (which could be using a different interface). This has been
210 changed to use the value of the helo_data option from the smtp transport
211 instead - this is what is used when a message is actually being sent. If
212 there is no remote transport (possible with a router that sets up host
213 addresses), $smtp_active_hostname is used.
215 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
216 tweaks were necessary in order to get it to work (see also 21 below):
217 (a) The code assumed that strncpy() returns a negative number on buffer
218 overflow, which isn't the case. Replaced with Exim's string_format()
220 (b) There were several signed/unsigned issues. I just did the minimum
221 hacking in of casts. There is scope for a larger refactoring.
222 (c) The code used strcasecmp() which is not a standard C function.
223 Replaced with Exim's strcmpic() function.
224 (d) The code set only $1; it now sets $auth1 as well.
225 (e) A simple test gave the error "authentication client didn't specify
226 service in request". It would seem that Dovecot has changed its
227 interface. Fortunately there's a specification; I followed it and
228 changed what the client sends and it appears to be working now.
230 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
233 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
234 address A is aliased to B and C, where B exists and C does not. Without
235 -v the output is "A verified" because verification stops after a
236 successful redirection if more than one address is generated. However,
237 with -v the child addresses are also verified. Exim was outputting "A
238 failed to verify" and then showing the successful verification for C,
239 with its parentage. It now outputs "B failed to verify", showing B's
240 parentage before showing the successful verification of C.
242 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
243 look up a TXT record in a specific list after matching in a combined
246 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
247 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
248 they consult the DNS. I had assumed they would set it the way they
249 wanted; and indeed my experiments on Linux seem to show that in some
250 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
251 To be on the safe side, however, I have now made the interface to
252 host_find_byname() similar to host_find_bydns(), with an argument
253 containing the DNS resolver options. The host_find_byname() function now
254 sets these options at its start, just as host_find_bydns() does. The smtp
255 transport options dns_qualify_single and dns_search_parents are passed to
256 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
257 of host_find_byname() use the default settings of RES_DEFNAMES
258 (qualify_single) but not RES_DNSRCH (search_parents).
260 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
261 spool_read_header() do less string testing, by means of a preliminary
262 switch on the second character of optional "-foo" lines. (This is
263 overdue, caused by the large number of possibilities that now exist.
264 Originally there were few.) While I was there, I also converted the
265 str(n)cmp tests so they don't re-test the leading "-" and the first
266 character, in the hope this might squeeze out yet more improvement.
268 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
269 flag allowing group syntax was set by the header_syntax check but not
270 turned off, possible causing trouble later; (2) The flag was not being
271 set at all for the header_verify test, causing "group"-style headers to
272 be rejected. I have now set it in this case, and also caused header_
273 verify to ignore an empty address taken from a group. While doing this, I
274 came across some other cases where the code for allowing group syntax
275 while scanning a header line wasn't quite right (mostly, not resetting
276 the flag correctly in the right place). These bugs could have caused
277 trouble for malformed header lines. I hope it is now all correct.
279 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
280 with the "reply" argument non-NULL. The code, however (which originally
281 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
282 but it didn't always do it. This confused somebody who was copying the
283 code for some other use. I have removed all the tests.
285 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
286 feature that was used to support insecure browsers during the U.S. crypto
287 embargo. It requires special client support, and Exim is probably the
288 only MTA that supported it -- and would never use it because real RSA is
289 always available. This code has been removed, because it had the bad
290 effect of slowing Exim down by computing (never used) parameters for the
291 RSA_EXPORT functionality.
293 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
294 authenticator to fail if there's a tab character in the incoming data
295 (there should never be unless someone is messing about, as it's supposed
296 to be base64-encoded). Also added, on Timo's advice, the "secured" option
297 if the connection is using TLS or if the remote IP is the same as the
298 local IP, and the "valid-client-cert option" if a client certificate has
301 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
302 authenticators. This can be used for authorization after authentication
303 succeeds. (In the case of plaintext, it servers for both authentication
306 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
307 if any retry times were supplied.
309 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
310 connection, where there is no client IP address to check. In this
311 situation, the verify now always succeeds.
313 PH/25 Applied John Jetmore's -Mset patch.
315 PH/26 Added -bem to be like -Mset, but loading a message from a file.
317 PH/27 In a string expansion for a processed (not raw) header when multiple
318 headers of the same name were present, leading whitespace was being
319 removed from all of them, but trailing whitespace was being removed only
320 from the last one. Now trailing whitespace is removed from each header
321 before concatenation. Completely empty headers in a concatenation (as
324 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
325 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
327 PH/29 [Removed. This was a change that I later backed out, and forgot to
328 correct the ChangeLog entry (that I had efficiently created) before
329 committing the later change.]
331 PH/30 Exim was sometimes attempting to deliver messages that had suffered
332 address errors (4xx response to RCPT) over the same connection as other
333 messages routed to the same hosts. Such deliveries are always "forced",
334 so retry times are not inspected. This resulted in far too many retries
335 for the affected addresses. The effect occurred only when there were more
336 hosts than the hosts_max_try setting in the smtp transport when it had
337 the 4xx errors. Those hosts that it had tried were not added to the list
338 of hosts for which the message was waiting, so if all were tried, there
339 was no problem. Two fixes have been applied:
341 (i) If there are any address or message errors in an SMTP delivery, none
342 of the hosts (tried or untried) are now added to the list of hosts
343 for which the message is waiting, so the message should not be a
344 candidate for sending over the same connection that was used for a
345 successful delivery of some other message. This seems entirely
346 reasonable: after all the message is NOT "waiting for some host".
347 This is so "obvious" that I'm not sure why it wasn't done
348 previously. Hope I haven't missed anything, but it can't do any
349 harm, as the worst effect is to miss an optimization.
351 (ii) If, despite (i), such a delivery is accidentally attempted, the
352 routing retry time is respected, so at least it doesn't keep
353 hammering the server.
355 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
356 in ${readsocket because some servers need this prod.
358 PH/32 Added some extra debug output when updating a wait-xxx database.
360 PH/33 The hint "could be header name not terminated by colon", which has been
361 given for certain expansion errors for a long time, was not being given
362 for the ${if def:h_colon_omitted{... case.
364 PH/34 The spec says: "With one important exception, whenever a domain list is
365 being scanned, $domain contains the subject domain." There was at least
366 one case where this was not true.
368 PH/35 The error "getsockname() failed: connection reset by peer" was being
369 written to the panic log as well as the main log, but it isn't really
370 panic-worthy as it just means the connection died rather early on. I have
371 removed the panic log writing for the ECONNRESET error when getsockname()
374 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
375 runs only) independently of the message's sender address. This meant
376 that, if the 4xx error was in fact related to the sender, a different
377 message to the same recipient with a different sender could confuse
378 things. In particualar, this can happen when sending to a greylisting
379 server, but other circumstances could also provoke similar problems.
380 I have changed the default so that the retry time for these errors is now
381 based a combination of the sender and recipient addresses. This change
382 can be overridden by setting address_retry_include_sender=false in the
385 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
386 remote server are returned as part of bounce messages. This was not
387 happening for LMTP over a pipe (the lmtp transport), but now it is the
388 same for both kinds of LMTP.
390 PH/38 Despite being documented as not happening, Exim was rewriting addresses
391 in header lines that were in fact CNAMEs. This is no longer the case.
393 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
394 and queue runs started by the daemon processed all messages. This has
395 been fixed so that -R and -S can now usefully be given with -q<time>.
397 PH/40 Import PCRE release 6.7 (fixes some bugs).
399 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
401 PH/42 Give an error if -q is specified more than once.
403 PH/43 Renamed the variables $interface_address and $interface_port as
404 $received_ip_address and $received_port, to make it clear that these
405 values apply to message reception, and not to the outgoing interface when
406 a message is delivered. (The old names remain recognized, of course.)
408 PH/44 There was no timeout on the connect() call when using a Unix domain
409 socket in the ${readsocket expansion. There now is.
411 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
412 be meaningful with "accept".
414 SC/01 Eximstats V1.43
415 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
417 SC/02 Eximstats V1.44
418 Use a glob alias rather than an array ref in the generated
419 parser. This improves both readability and performance.
421 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
422 Collect SpamAssassin and rejection statistics.
423 Don't display local sender or destination tables unless
424 there is data to show.
425 Added average volumes into the top table text output.
427 SC/04 Eximstats V1.46
428 Collect data on the number of addresses (recipients)
429 as well as the number of messages.
431 SC/05 Eximstats V1.47
432 Added 'Message too big' to the list of mail rejection
433 reasons (thanks to Marco Gaiarin).
435 SC/06 Eximstats V1.48
436 Mainlog lines which have GMT offsets and are too short to
437 have a flag are now skipped.
439 SC/07 Eximstats V1.49 (Alain Williams)
440 Added the -emptyok flag.
442 SC/08 Eximstats V1.50
443 Fixes for obtaining the IP address from reject messages.
445 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
446 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
447 whitesspace changes from 4.64-PH/27
449 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
452 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
453 are found, allow negative numbers in numeric criteria)
455 JJ/06 exipick.20061117.2, added new $message_body_missing variable
457 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
458 to match changes made in 4.64-PH/43
460 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
462 PH/47 Put in an explicit test for a DNS lookup of an address record where the
463 "domain" is actually an IP address, and force a failure. This locks out
464 those revolvers/nameservers that support "A-for-A" lookups, in
465 contravention of the specifications.
467 PH/48 When a host name was looked up from an IP address, and the subsequent
468 forward lookup of the name timed out, the host name was left in
469 $sender_host_name, contrary to the specification.
471 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
472 restricted to single-key lookups, Exim was not diagnosing an error if
473 * or *@ was used with a query-style lookup.
475 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
477 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
478 long ago, but noone interested enough thought of it. Let's just say that
479 the "1.1" means that there are some new functions that weren't there at
480 some point in the past.
482 PH/51 Error processing for expansion failure of helo_data from an smtp
483 transport during callout processing was broken.
485 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
486 tested/used via the -bh/-bhc/-bs options.
488 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
489 bug, fixed in subsequent PCRE releases).
491 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
492 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
494 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
500 SC/01 Use a glob alias rather than an array ref in eximstats generated
501 parser. This improves both readability and performance.
503 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
504 Don't display local sender or destination tables in eximstats unless
505 there is data to show.
506 Added average volumes into the eximstats top table text output.
508 SC/03 Collect data on the number of addresses (recipients) as well
509 as the number of messages in eximstats.
511 TF/01 Correct an error in the documentation for the redirect router. Exim
512 does (usually) call initgroups() when daemonizing.
514 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
515 with consistent privilege compared to when running as a daemon.
517 TF/03 Note in the spec that $authenticated_id is not set for local
518 submissions from trusted users.
520 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
521 Thanks to Dean Brooks <dean@iglou.com> for the patch.
523 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
524 by adding some example configuration directives to the default
525 configuration file. A little bit of work is required to uncomment the
526 directives and define how usernames and passwords are checked, but
527 there is now a framework to start from.
529 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
530 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
531 without this. I don't know how relevant this is to other LDAP libraries.
533 PH/02 Add the verb name to the "unknown ACL verb" error.
535 PH/03 Magnus Holmgren's patch for filter_prepend_home.
537 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
539 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
540 directory not expanded when it should be if an expanded home directory
541 was set for the address (which is overridden by the transport).
543 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
546 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
547 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
548 because it is too late at that time, and has no effect.
550 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
551 security issue with \' (bugzilla #107). I could not use the
552 PQescapeStringConn() function, because it needs a PGconn value as one of
555 PH/08 When testing addresses using -bt, indicate those final addresses that
556 are duplicates that would not cause an additional delivery. At least one
557 person was confused, thinking that -bt output corresponded to deliveries.
558 (Suppressing duplicates isn't a good idea as you lose the information
559 about possibly different redirections that led to the duplicates.)
561 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
562 systems where poll() doesn't work, in particular OS X.
564 PH/10 Added more information to debugging output for retry time not reached.
566 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
567 operations in malware.c.
569 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
572 PH/13 If write_rejectlog was set false when logging was sent to syslog with
573 syslog_duplication set false, log lines that would normally be written
574 both the the main log and to the reject log were not written to syslog at
577 PH/14 In the default configuration, change the use of "message" in ACL warn
578 statements to "add_header".
580 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
581 not followed by a command (e.g. "seen endif").
583 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
584 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
587 PH/17 Added extra conditions to the default value of delay_warning_condition
591 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
592 { match{$h_precedence:}{(?i)bulk|list|junk} } \
593 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
596 The Auto-Submitted: and various List- headers are standardised, whereas I
597 don't think Precedence: ever was.
599 PH/18 Refactored debugging code in route_finduser() to show more information,
600 in particular, the error code if getpwnam() issues one.
602 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
603 This is apparently needed in addition to the PH/07 change above to avoid
604 any possible encoding problems.
606 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
607 but not after initializing Perl.
609 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
610 output them only if debugging. By default they are written stderr,
611 apparently, which is not desirable.
613 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
616 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
619 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
621 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
622 authenticated or an ident call has been made. Suppress the default
623 values for $authenticated_id and $authenticated_sender (but permit -oMai
624 and -oMas) when testing with -bh.
626 PH/24 Re-jigged the order of the tests in the default configuration so that the
627 tests for valid domains and recipients precede the DNS black list and CSA
628 tests, on the grounds that those ones are more expensive.
630 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
631 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
632 This bug exists in every version of Exim that I still have, right back to
635 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
636 However, an attempt to turn on case-sensitivity in a regex key by
637 including (?-i) didn't work because the subject string was already
638 lowercased, and the effects were non-intuitive. It turns out that a
639 one-line patch can be used to allow (?-i) to work as expected.
645 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
646 other effects) broke the use of negated acl sub-conditions.
648 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
651 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
652 "Deny" causes Exim to reject the incoming connection with a 554 error.
653 Unfortunately, if there is a major crisis, such as a disk failure,
654 tcp-wrappers gives "deny", whereas what one would like would be some
655 kind of temporary error. A kludge has been added to help with this.
656 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
657 554 error is used if errno is still zero or contains ENOENT (which occurs
658 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
661 PH/03 Add -lutil to the default FreeBSD LIBS setting.
663 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
664 errors. Otherwise a message that provokes a temporary error (when other
665 messages do not) can cause a whole host to time out.
667 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
668 addresses were routed directly to files or pipes from a redirect router.
669 File deliveries just didn't batch; pipe deliveries might have suffered
672 PH/06 A failure to get a lock for a hints database would erroneously always say
673 "Failed to get write lock", even when it was really a read lock.
675 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
676 of 0600. This has been changed to use the value of the lockfile_mode
677 option (which defaults to 0600).
679 PH/08 Applied small patch from the Sieve maintainer.
681 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
682 folder from quota calculations, a direct delivery into this folder messed
683 up the contents of the maildirsize file. This was because the regex was
684 used only to exclude .Trash (or whatever) when the size of the mailbox
685 was calculated. There was no check that a delivery was happening into an
686 excluded directory. This bug has been fixed by ignoring all quota
687 processing for deliveries into excluded directories.
689 PH/10 Added the maildirfolder_create_regex option to appendfile.
695 PH/01 The code for finding all the local interface addresses on a FreeBSD
696 system running IPv6 was broken. This may well have applied to all BSD
697 systems, as well as to others that have similar system calls. The broken
698 code found IPv4 interfaces correctly, but gave incorrect values for the
699 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
700 that it would not match correctly against @[] and not recognize the IPv6
703 PH/02 The ipliteral router was not recognizing addresses of the form user@
704 [ipv6:....] because it didn't know about the "ipv6:" prefix.
706 PH/03 Added disable_ipv6.
708 PH/04 Changed $reply_address to use the raw form of the headers instead of the
709 decoded form, because it is most often used to construct To: headers
710 lines in autoreplies, and the decoded form may well be syntactically
711 invalid. However, $reply_address has leading white space removed, and all
712 newlines turned into spaces so that the autoreply transport does not
715 PH/05 If group was specified without a user on a router, and no group or user
716 was specified on a transport, the group from the router was ignored.
718 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
719 for visible compile-time settings that can be used to change these
720 numbers, for those that want even more. Backwards compatibility with old
721 spool files has been maintained. However, going back to a previous Exim
722 release will lost any variables that are in spool files.
724 PH/07 Two small changes when running in the test harness: increase delay when
725 passing a TCP/IP connection to a new process, in case the original
726 process has to generate a bounce, and remove special handling of
727 127.0.0.2 (sic), which is no longer necessary.
729 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
730 be the same on different OS.
732 PH/09 Moved a debug statement in filter processing to avoid a race problem when
735 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
736 whether --show-vars was specified or not
738 JJ/02 exipick: Added support for new ACL variable spool format introduced
741 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
742 syntactically invalid From: or Reply-to: line, and a filter used this to
743 generate an autoreply, and therefore failed to obtain an address for the
744 autoreply, Exim could try to deliver to a non-existent relative file
745 name, causing unrelated and misleading errors. What now happens is that
746 it logs this as a hard delivery error, but does not attempt to create a
749 PH/11 The exinext utility has a -C option for testing purposes, but although
750 the given file was scanned by exinext itself; it wasn't being passed on
751 when Exim was called.
753 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
754 an end-of-file indication when reading a command response.
756 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
757 compiled. In many other places in Exim, IPv6 addresses are always
758 recognized, so I have changed this. It also means that IPv4 domain
759 literals of the form [IPV4:n.n.n.n] are now always recognized.
761 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
762 used if the router is not running as root, for example, when verifying at
763 ACL time, or when using -bh. The debugging output from this situation was
764 non-existent - all you got was a failure to exec. I have made two
767 (a) Failures to set uid/gid, the current directory, or a process leader
768 in a subprocess such as that created by queryprogram now generate
769 suitable debugging ouput when -d is set.
771 (b) The queryprogram router detects when it is not running as root,
772 outputs suitable debugging information if -d is set, and then runs
773 the subprocess without attempting to change uid/gid.
775 PH/15 Minor change to Makefile for building test_host (undocumented testing
778 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
779 additional section of a DNS packet that returns MX or SRV records.
780 Instead, it always explicitly searches for A/AAAA records. This avoids
781 major problems that occur when a DNS server includes only records of one
782 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
783 fixed another bug: if SRV records were looked up and the corresponding
784 address records were *not* found in the additional section, the port
785 values from the SRV records were lost.
787 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
788 using the correct key (the original address) when searching the retry
789 rules in order to find which one to use for generating the retry hint.
791 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
792 adding the default one. Similarly, if it contains a Reply-To: header, the
793 errors_reply_to option, if set, is not used.
795 PH/19 When calculating a retry time, Exim used to measure the "time since
796 failure" by looking at the "first failed" field in the retry record. Now
797 it does not use this if it is later than than the arrival time of the
798 message. Instead it uses the arrival time. This makes for better
799 behaviour in cases where some deliveries succeed, thus re-setting the
800 "first failed" field. An example is a quota failure for a huge message
801 when small messages continue to be delivered. Without this change, the
802 "time since failure" will always be short, possible causing more frequent
803 delivery attempts for the huge message than are intended.
804 [Note: This change was subsequently modified - see PH/04 for 4.62.]
806 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
807 $1, $2, $3) because the numerical variables can be reset during some
808 expansion items (e.g. "match"), thereby losing the authentication data.
810 PH/21 Make -bV show the size of off_t variables so that the test suite can
811 decide whether to run tests for quotas > 2G.
813 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
814 mailbox_size, and mailbox_filecount in the appendfile transport. If a
815 filecount value is greater than 2G or if a quota value is greater than 2G
816 on a system where the size of off_t is not greater than 4, a panic error
819 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
820 never match. The debug and -bh output now contains an explicit error
821 message indicating a malformed IPv4 address or mask.
823 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
824 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
827 PH/25 Do not write to syslog when running in the test harness. The only
828 occasion when this arises is a failure to open the main or panic logs
829 (for which there is an explicit test).
831 PH/26 Added the /no_tell option to "control=freeze".
833 PH/27 If a host name lookup failed very early in a connection, for example, if
834 the IP address matched host_lookup and the reverse lookup yielded a name
835 that did not have a forward lookup, an error message of the form "no IP
836 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
837 could be logged. Now it outputs the IP address instead of "NULL".
839 PH/28 An enabling patch from MH: add new function child_open_exim2() which
840 allows the sender and the authenticated sender to be set when
841 submitting a message from within Exim. Since child_open_exim() is
842 documented for local_scan(), the new function should be too.
844 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
845 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
846 results in an empty string is now treated as unset.
848 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
850 PH/31 Added sender_verify_fail logging option.
852 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
853 needed by RFC 822 but not by RFC 2822 was commented out. I have now
854 tidied the source and removed it altogether.
856 PH/33 When a queue run was abandoned because the load average was too high, a
857 log line was always written; now it is written only if the queue_run log
858 selector is set. In addition, the log line for abandonment now contains
859 information about the queue run such as the pid. This is always present
860 in "start" and "stop" lines but was omitted from the "abandon" line.
862 PH/34 Omit spaces between a header name and the colon in the error message that
863 is given when verify = headers_syntax fails (if there are lots of them,
864 the message gets confusing).
866 PH/35 Change the default for dns_check_names_pattern to allow slashes within
867 names, as there are now some PTR records that contain slashes. This check
868 is only to protect against broken name servers that fall over on strange
869 characters, so the fact that it applies to all lookups doesn't matter.
871 PH/36 Now that the new test suite is complete, we can remove some of the
872 special code in Exim that was needed for the old test suite. For example,
873 sorting DNS records because real resolvers return them in an arbitrary
874 order. The new test suite's fake resolver always returns records in the
877 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
878 bounces) except when queue_only is set, to avoid logging races between
879 the different processes.
881 PH/38 Panic-die if .include specifies a non-absolute path.
883 PH/39 A tweak to the "H" retry rule from its user.
885 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
886 a label. They prevented compilation on older perls.
888 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
889 a warning to be raised on newish perls.
891 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
892 on queue. Changes to match documented behaviour of showing count of
893 messages matching specified criteria.
895 PH/40 Changed the default ident timeout from 30s to 5s.
897 PH/41 Added support for the use of login_cap features, on those BSD systems
898 that have them, for controlling the resources used by pipe deliveries.
900 PH/42 The content-scanning code uses fopen() to create files in which to put
901 message data. Previously it was not paying any attention to the mode of
902 the files. Exim runs with umask(0) because the rest of the code creates
903 files with open(), and sets the required mode explicitly. Thus, these
904 files were ending up world-writeable. This was not a big issue, because,
905 being within the spool directory, they were not world-accessible. I have
906 created a function called modefopen, which takes an additional mode
907 argument. It sets umask(777), creates the file, chmods it to the required
908 mode, then resets the umask. All the relevant calls to fopen() in the
909 content scanning code have been changed to use this function.
911 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
912 to 24 hours. This avoids potential overflow problems when processing G
913 and H retry rules. I suspect nobody ever tinkers with this value.
915 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
917 PH/45 When the plaintext authenticator is running as a client, the server's
918 challenges are checked to ensure they are valid base64 strings. By
919 default, the authentication attempt is cancelled if an invalid string is
920 received. Setting client_ignore_invalid_base64 true ignores these errors.
921 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
922 they are received. Thus, the responses can be made to depend on the
923 challenges. If an invalid string is ignored, an empty string is placed in
926 PH/46 Messages that are created by the autoreply transport now contains a
927 References: header, in accordance with RFCs 2822 and 3834.
929 PH/47 Added authenticated_sender_force to the smtp transport.
931 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
933 PH/49 Installed latest patch from the Sieve maintainer.
935 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
936 was also set, the appendfile transport was unnecessarily scanning a
937 directory of message files (e.g. for maildir delivery) to find the count
938 of files (along with the size), even though it did not need this
939 information. It now does the scan only if it needs to find either the
940 size of the count of files.
942 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
944 PH/52 Two bugs concerned with error handling when the smtp transport is
947 (i) Exim was not creating retry information for temporary errors given
948 for individual recipients after the DATA command when the smtp transport
949 was used in LMTP mode. This meant that they could be retried too
950 frequently, and not timed out correctly.
952 (ii) Exim was setting the flag that allows error details to be returned
953 for LMTP errors on RCPT commands, but not for LMTP errors for individual
954 recipients that were returned after the DATA command.
956 PH/53 This is related to PH/52, but is more general: for any failing address,
957 when detailed error information was permitted to be returned to the
958 sender, but the error was temporary, then after the final timeout, only
959 "retry timeout exceeded" was returned. Now it returns the full error as
960 well as "retry timeout exceeded".
962 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
963 do this, and (what is worse) MTAs that accept it.
965 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
966 will now be deprecated.
968 PH/56 New os.c-cygwin from the Cygwin maintainer.
970 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
971 formats (previously only available in exim formats via -bpr, -bpru,
972 and -bpra. Now also available in native and exiqgrep formats)
974 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
975 with very large, slow to parse queues
977 JJ/08 exipick: added ! as generic prefix to negate any criteria format
979 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
981 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
982 responses to authentication challenges, though it was showing the
983 challenges; (ii) I've removed the CR characters from the debug output for
986 PH/58 Allow for the insertion of a newline as well as a space when a string
987 is turned into more than one encoded-word during RFC 2047 encoding. The
988 Sieve code now uses this.
990 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
991 data_4xx, lost_connection, tls_required.
993 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
994 message was being sent as an SMTP response.
996 PH/61 Add -l and -k options to exicyclog.
998 PH/62 When verifying, if an address was redirected to one new address, so that
999 verification continued, and the new address failed or deferred after
1000 having set something in $address_data, the value of $address_data was not
1001 passed back to the ACL. This was different to the case when no
1002 redirection occurred. The value is now passed back in both cases.
1004 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1005 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1006 use login_cap.h, so on its own it isn't the distinguishing feature. The
1007 new name refers directly to the setclassresources() function.
1009 PH/65 Added configuration files for NetBSD3.
1011 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1013 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1014 is preferred over IPv4.
1016 PH/68 The bounce_return_message and bounce_return_body options were not being
1017 honoured for bounces generated during the reception of non-SMTP messages.
1018 In particular, this applied to messages rejected by the ACL. This bug has
1019 been fixed. However, if bounce_return_message is true and bounce_return_
1020 body is false, the headers that are returned for a non-SMTP message
1021 include only those that have been read before the error was detected.
1022 (In the case of an ACL rejection, they have all been read.)
1024 PH/69 The HTML version of the specification is now built in a directory called
1025 spec_html instead of spec.html, because the latter looks like a path with
1026 a MIME-type, and this confuses some software.
1028 PH/70 Catch two compiler warnings in sieve.c.
1030 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1031 function verify_get_ident() calls ip_connect() to connect a socket, but
1032 if the "connect()" function timed out, ip_connect() used to close the
1033 socket. However, verify_get_ident() also closes the socket later, and in
1034 between Exim writes to the log, which may get opened at this point. When
1035 the socket was closed in ip_connect(), the log could get the same file
1036 descriptor number as the socket. This naturally causes chaos. The fix is
1037 not to close the socket in ip_connect(); the socket should be closed by
1038 the function that creates it. There was only one place in the code where
1039 this was missing, in the iplookup router, which I don't think anybody now
1040 uses, but I've fixed it anyway.
1042 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1043 well as to direct DNS lookups. Otherwise the handling of names in host
1044 lists is inconsistent and therefore confusing.
1050 PH/01 Two changes to the default runtime configuration:
1052 (1) Move the checks for relay_from_hosts and authenticated clients from
1053 after to before the (commented out) DNS black list checks.
1055 (2) Add control=submission to the relay_from_hosts and authenticated
1056 clients checks, on the grounds that messages accepted by these
1057 statements are most likely to be submissions.
1059 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1061 (1) Generate an error if the third argument for the ${prvs expansion is
1064 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1067 (3) Reset the variables that are obtained from the first argument of
1068 ${prvscheck and used in the second argument before leaving the code,
1069 because their memory is reclaimed, so using them afterwards may do
1072 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1073 one (it's much easier than Tom thought :-).
1075 (5) Because of (4), we can now allow for the use of $prvscheck_result
1076 inside the third argument.
1078 PH/03 For some reason, the default setting of PATH when running a command from
1079 a pipe transport was just "/usr/bin". I have changed it to
1082 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1083 anything to be listed in the output from -bV.
1085 PH/05 When a filter generated an autoreply, the entire To: header line was
1086 quoted in the delivery log line, like this:
1088 => >A.N.Other <ano@some.domain> <original@ddress> ...
1090 This has been changed so that it extracts the operative address. There
1091 may be more than one such address. If so, they are comma-separated, like
1094 => >ano@some.domain,ona@other.domain <original@ddress> ...
1096 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1097 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1098 not being looked up in the rDNS to get a host name, Exim was showing the
1099 IP address twice in Received: lines, even though the IP addresses were
1100 identical. For example:
1102 Received: from [1.2.3.4] (helo=[1.2.3.4])
1104 However, if the real host name was known, it was omitting the HELO data
1105 if it matched the actual IP address. This has been tidied up so that it
1106 doesn't show the same IP address twice.
1108 PH/07 When both +timestamp and +memory debugging was on, the value given by
1109 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1110 called by the debug printing, thereby overwriting the timestamp buffer.
1111 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1113 PH/08 When the original message was included in an autoreply transport, it
1114 always said "this is a copy of the message, including all the headers",
1115 even if body_only or headers_only was set. It now gives an appropriate
1118 PH/09 Applied a patch from the Sieve maintainer which:
1120 o fixes some comments
1121 o adds the (disabled) notify extension core
1122 o adds some debug output for the result of if/elsif tests
1123 o points to the current vacation draft in the documentation
1124 and documents the missing references header update
1128 o fixes a bug in processing the envelope test (when testing
1129 multiple envelope elements, the last element determinted the
1132 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1133 Electronic Mail") by including:
1135 Auto-submitted: auto-generated
1137 in the messages that it generates (bounce messages and others, such as
1138 warnings). In the case of bounce messages for non-SMTP mesages, there was
1139 also a typo: it was using "Auto_submitted" (underscore instead of
1140 hyphen). Since every message generated by Exim is necessarily in response
1141 to another message, thes have all been changed to:
1143 Auto-Submitted: auto-replied
1145 in accordance with these statements in the RFC:
1147 The auto-replied keyword:
1149 - SHOULD be used on messages sent in direct response to another
1150 message by an automatic process,
1152 - MUST NOT be used on manually-generated messages,
1154 - MAY be used on Delivery Status Notifications (DSNs) and Message
1155 Disposition Notifications (MDNs),
1157 - MUST NOT be used on messages generated by automatic or periodic
1158 processes, except for messages which are automatic responses to
1161 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1162 to the default Received: header definition.
1164 PH/12 Added log selector acl_warn_skipped (default on).
1166 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1167 variables because (a) they are in the wrong storage pool and (b) even if
1168 they were copied, it wouldn't work properly because of the caching.
1170 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1171 checking when decoding. Apparently there are clients that generate
1172 overlong encoded strings. Why am I not surprised?
1174 PH/15 If the first argument of "${if match_address" was not empty, but did not
1175 contain an "@" character, Exim crashed. Now it writes a panic log message
1176 and treats the condition as false.
1178 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1180 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1181 extension "envelope-auth". The code is finished and in agreement with
1182 other implementations, but there is no documentation so far and in fact,
1183 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1184 not changing the active code.
1186 Print executed "if" and "elsif" statements when debugging is used. This
1187 helps a great deal to understand what a filter does.
1189 Document more things not specified clearly in RFC3028. I had all this
1190 sorted out, when out of a sudden new issues came to my mind. Oops."
1192 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1195 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1196 canonical form (as documented). However, after a host name lookup from
1197 the IP address, check_host() was doing a simple string comparison with
1198 addresses acquired from the DNS when checking that the found name did
1199 have the original IP as one of its addresses. Since any found IPv6
1200 addresses are likely to be in abbreviated form, the comparison could
1201 fail. Luckily, there already exists a function for doing the comparison
1202 by converting both addresses to binary, so now that is used instead of
1203 the text comparison.
1205 PH/20 There was another similar case to PH/19, when a complete host name was
1206 given in a host list; looking up its IP address could give an abbreviated
1207 form, whereas the current host's name might or might not be abbreviated.
1208 The same fix has been applied.
1214 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1215 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1218 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1219 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1221 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1223 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1224 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1225 Also, for "Auto-Submitted", treat anything other than "no" as
1226 non-personal, in accordance with RFC 3834. (Previously it treated
1227 anything starting "auto-" as non-personal.)
1229 TF/01 The control=submission/name=... option had a problem with syntax
1230 errors if the name included a slash character. The /name= option
1231 now slurps the rest of the string, so it can include any characters
1232 but it must come last in the list of options (after /sender_retain
1235 PH/05 Some modifications to the interface to the fake nameserver for the new
1243 TK/01 Added the "success_on_redirect" address verification option. See
1244 NewStuff for rationale and an example.
1246 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1248 PH/02 Patch to exigrep to allow it to work on syslog lines.
1250 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1251 fread() to skip over the body file's header line, because in Cygwin the
1252 header line is locked and is inaccessible.
1254 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1255 co-exist for some time) to make it clear that it is the Exim ID that is
1256 referenced, not the Message-ID: header line.
1258 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1259 string_format() function, because snprintf() does not exist on all
1262 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1263 ${stat: expansion item.
1265 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1266 protocol synchronization error", to keep the pedants happy.
1268 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1269 well as for IRIX systems, when gcc is being used. See the host.c source
1272 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1274 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1277 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1278 transport and to the smtp transport in LMTP mode.
1280 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1282 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1283 run a filter in a subprocess. This could lead to confusion in subsequent
1284 lookups in the parent process. There should also be a search_tidyup() at
1285 the end of the subprocess.
1287 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1288 only if the host matched helo_try_verify_hosts, which caused the
1289 verification to occur when the EHLO/HELO command was issued. The ACL just
1290 tested the remembered result. Now, if a previous verification attempt has
1291 not happened, "verify = helo" does it there and then.
1293 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1295 TK/03 Fix log output including CR from clamd.
1297 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1298 exist provoked a memory error which could cause a segfault.
1300 PH/15 Installed PCRE 6.2
1302 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1304 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1305 of the problem. Specifically, suggested +O2 rather than +O1 for the
1308 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1310 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1311 record, the port was not correctly logged when the outgoing_port log
1312 selector was set (it logged the transort's default port).
1314 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1315 fallback_hosts, and "hosts" in the smtp transport.
1317 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1318 host errors such as "Connection refused".
1320 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1321 authentication with radiusclient 0.4.9:
1323 - Error returned from rc_read_config was caught wrongly
1324 - Username/password not passed on to radius server due to wrong length.
1326 The presumption is that some radiusclient API changes for 4.51/PH/17
1327 were not taken care of correctly. The code is still untested by me (my
1328 Linux distribution still has 0.3.2 of radiusclient), but it was
1329 contributed by a Radius user.
1331 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1332 expanding the "port" option of the smtp transport.
1334 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1335 while reading a MIME header. Thanks to Tom Hughes for a patch.
1337 PH/24 Include config.h inside local_scan.h so that configuration settings are
1340 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1341 that in an ACL for RCPT (for example), you can examine exactly what was
1344 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1345 commands, but it was not correctly comparing the address with the actual
1346 client host address. Thus, it would show the EHLO address in Received:
1347 header lines when this was not necessary.
1349 PH/27 Added the % operator to ${eval:}.
1351 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1352 it should be ignoring failures (because with -C, for example, it has lost
1353 privilege). It wasn't ignoring creation failures other than "already
1356 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1359 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1360 because an input line was too long, either on its own, or by virtue of
1361 too many continuations, the temporary file was not being removed, and the
1362 return code was incorrect.
1364 PH/31 Missing "BOOL" in function definition in filtertest.c.
1366 PH/32 Applied Sieve patches from the maintainer.
1368 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1370 PH/33 Added "verify = not_blind".
1372 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1373 Local/Makefile (with some defaults set). These are used in built scripts
1374 such as exicyclog, but they have never been used in the exim_install
1375 script (though there are many overriding facilities there). I have
1376 arranged that the exim_install script now takes note of these two
1379 PH/35 Installed configuration files for Dragonfly.
1381 PH/36 When a locally submitted message by a trusted user did not contain a
1382 From: header, and the sender address was obtained from -f or from an SMTP
1383 MAIL command, and the trusted user did not use -F to supply a sender
1384 name, $originator_name was incorrectly used when constructing a From:
1385 header. Furthermore, $originator_name was used for submission mode
1386 messages from external hosts without From: headers in a similar way,
1387 which is clearly wrong.
1389 PH/37 Added control=suppress_local_fixups.
1391 PH/38 When log_selector = +received_sender was set, and the addition of the
1392 sender made the log line's construction buffer exactly full, or one byte
1393 less than full, an overflow happened when the terminating "\n" was
1396 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1397 when the result of a list match is failure because a DNS lookup failed.
1399 PH/40 RM_COMMAND is now used in the building process.
1401 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1402 the "build-* directories that it finds.
1404 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1405 address in a domain literal was a prefix of an interface address.
1407 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1408 when verifying a sender address, unless rewrite_headers is false.
1410 PH/44 Wrote a long comment about why errors_to addresses are verified as
1411 recipients, not senders.
1413 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1414 the ratelimit ACL was added.
1416 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1418 PH/46 Added extra information about PostgreSQL errors to the error string.
1420 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1421 avoiding the need to install special zones in a real server. This is
1422 backwards compatible; if it can't find the fake resolver, it drops back.
1423 Thus, both old and new test suites can be run.
1425 TF/02 Added util/ratelimit.pl
1427 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1428 clock is set back in time.
1430 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1431 Candler <B.Candler@pobox.com>.
1433 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1434 OK for addresses that are the result of redirections.
1436 PH/48 A number of further additions for the benefit of the new test suite,
1437 including a fake gethostbyname() that interfaces to the fake DNS resolver
1440 TF/06 The fix for widen_domains has also been applied to qualify_single and
1441 search_parents which are the other dnslookup options that can cause
1444 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1447 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1449 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1450 service" if it encountered an MX record with an empty target hostname.
1451 The message is now "an MX or SRV record indicated no SMTP service".
1453 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1454 if the DNS of the sending site is misconfigured. This is quite a
1455 common situation. This change restores the behaviour of treating a
1456 helo verification defer as a failure.
1458 PH/51 If self=fail was set on a router, the bounce message did not include the
1459 actual error message.
1465 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1467 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1468 command itself ended in error, the underlying message about the transport
1469 filter timeout was being overwritten with the pipe command error. Now the
1470 underlying error message should be appended to the second error message.
1472 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1474 PH/02 Reduce the amount of output that "make" produces by default. Full output
1475 can still be requested.
1477 PH/03 The warning log line about a condition test deferring for a "warn" verb
1478 was being output only once per connection, rather than after each
1479 occurrence (because it was using the same function as for successful
1480 "warn" verbs). This seems wrong, so I have changed it.
1482 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1483 it should not have, which might have caused a crash in the right
1484 circumstances, but probably never did.
1486 PH/04 Installed a modified version of Tony Finch's patch to make submission
1487 mode fix the return path as well as the Sender: header line, and to
1488 add a /name= option so that you can make the user's friendly name appear
1491 TF/03 Added the control = fakedefer ACL modifier.
1493 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1494 Mark Lowes for thorough testing.
1496 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1498 TK/03 Merged latest SRS patch from Miles Wilton.
1500 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1501 with the definition in sysexits.h (which is #included earlier).
1502 Fortunately, Exim does not actually use EX_OK. The code used to try to
1503 preserve the sysexits.h value, by assumimg that macro definitions were
1504 scanned for macro replacements. I have been disabused of this notion,
1505 so now the code just undefines EX_OK before #including unistd.h.
1507 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1508 in the smtp transport. When a block could not be written in a single
1509 write() function, the timeout was being re-applied to each part-write.
1510 This seems wrong - if the receiver was accepting one byte at a time it
1511 would take for ever. The timeout is now adjusted when this happens. It
1512 doesn't have to be particularly precise.
1514 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1515 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1517 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1518 without a domain if the check to <postmaster@domain> fails.
1520 SC/01 Eximstats: added -xls and the ability to specify output files
1521 (patch written by Frank Heydlauf).
1523 SC/02 Eximstats: use FileHandles for outputing results.
1525 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1527 SC/04 Eximstats: fixed display of large numbers with -nvr option
1529 SC/05 Eximstats: fixed merging of reports with empty tables.
1531 SC/06 Eximstats: added the -include_original_destination flag
1533 SC/07 Eximstats: removed tabs and trailing whitespace.
1535 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1537 TK/06 MBOX spool code: Add real "From " MBOX separator line
1538 so the .eml file is really in mbox format (even though
1539 most programs do not really care). Patch from Alex Miller.
1541 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1542 The latter is generated from $received_to and is only set if the
1543 message has one envelope recipient. SA can use these headers,
1544 obviously out-of-the-box. Patch from Alex Miller.
1546 PH/08 The ${def test on a variable was returning false if the variable's
1547 value was "0", contrary to what the specification has always said!
1548 The result should be true unless the variable is empty.
1550 PH/09 The syntax error of a character other than { following "${if
1551 def:variable_name" (after optional whitespace) was not being diagnosed.
1552 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1553 accidental colon was present, for example, could give incorrect results.
1555 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1556 result is used (not including appendfile, where other changes are about
1559 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1560 This involved changing a lot of size variables from int to off_t. It
1561 should work with maildirs and everything.
1563 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1564 spamd dying while we are connected to it.
1566 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1569 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1570 function for BDB 4.3.
1572 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1574 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1575 the sources for PCRE have been re-arranged and more files are now
1578 PH/15 The code I had for printing potentially long long variables in PH/11
1579 above was not the best (it lost precision). The length of off_t variables
1580 is now inspected at build time, and an appropriate printing format (%ld
1581 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1582 to be "long long int" or "long int". This is needed for the internal
1583 formatting function string_vformat().
1585 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1586 the configuration file to be ":syslog", then the script "guesses" where
1587 the logs files are, rather than using the compiled in default. In our
1588 case the guess is not the same as the compiled default, so the script
1589 suddenly stopped working when I started to use syslog. The patch checks
1590 to see if log_file_path is "". If so, it attempts to read it from exim
1591 with no configuration file to get the compiled in version, before it
1592 falls back to the previous guessing code."
1594 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1595 implementing BATV in an Exim configuration. See NewStuff for the gory
1598 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1599 Makefile that are specific to HP-UX.
1601 PH/18 If the "use_postmaster" option was set for a recipient callout together
1602 with the "random" option, the postmaster address was used as the MAIL
1603 FROM address for the random test, but not for the subsequent recipient
1604 test. It is now used for both.
1606 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1607 patch removes a few documentation additions to RFC 3028, because the
1608 latest draft now contains them. It adds the new en;ascii-case comparator
1609 and a new error check for 8bit text in MIME parts. Comparator and
1610 require names are now matched exactly. I enabled the subaddress
1611 extension, but it is not well tested yet (read: it works for me)."
1613 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1614 rework some of the code of TK/09 above to avoid the hardwired use of
1615 "%lld" and "long long". Replaced the call to snprintf() with a call to
1618 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1619 records point to non-existent hosts", "retry timeout exceeded", and
1620 "retry time not reached for any host after a long failure period".
1622 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1623 experimental DomainKeys support:
1625 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1626 (2) On an error such as an illegally used "control", the wrong name for
1627 the control was given.
1629 These problems did NOT occur unless DomainKeys support was compiled.
1631 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1633 PH/24 Added ${if match_ip condition.
1635 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1636 databases so that it will be absolutely obvious if a crash occurs in the
1637 DB library. This is a regular occurrence (often caused by mis-matched
1640 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1641 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1642 picked up on a user's system that detects such things. There doesn't seem
1643 to be a gcc warning option for this - only an attribute that has to be
1644 put on the function's prototype. It seems that in Fedora Core 4 they have
1645 set this on a number of new functions. No doubt there will be more in due
1648 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1649 specify a transport. However, if an address that was verified by such a
1650 router was the subject of a callout, Exim crashed because it tried to
1651 read the rcpt_include_affixes from the non-existent transport. Now it
1652 just assumes that the setting of that option is false. This bug was
1653 introduced by 4.51/PH/31.
1655 PH/28 Changed -d+all to exclude +memory, because that information is very
1656 rarely of interest, but it makes the output a lot bigger. People tend to
1657 do -d+all out of habit.
1659 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1660 code in os-type was giving problems when libc.so lives in lib64, like on
1663 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1664 aren't the modern standard, and it seems that some systems' include files
1665 don't always have them. Exim was already checking for some of the newer
1666 ones like T_AAAA, and defining it itself. I've added checks for all the
1667 record types that Exim uses.
1669 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1670 not automatically generating a new one, as it is supposed to. This
1671 prevented TLS from working. If the file did exist, but contained invalid
1672 data, a new version was generated, as expected. It was only the case of a
1673 non-existent file that was broken.
1675 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1676 with a change in libdomainkeys > 0.64.
1678 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1679 from DNS. If the selector record carries the flag, it now has
1680 precedence over the domain-wide flag.
1682 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1684 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1685 the use of a port name that isn't defined in /etc/services) occurred, the
1686 message was deferred as in a normal delivery, and thus remained on the
1687 spool, instead of being failed because of the mua_wrapper setting. This
1688 is now fixed, and I tidied up some of the mua_wrapper messages at the
1691 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1692 the messages in a hash of arrays rather than using individual hashes.
1693 This is a bit cleaner and results in dramatic memory savings, albeit
1694 at a slight CPU cost.
1696 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1697 as requested by Marc Sherman.
1699 SC/10 Eximstats: added histograms for user specified patterns as requested
1702 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1704 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1705 fopen() in the content-scanning modules that did not already have it.
1711 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1712 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1714 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1716 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1718 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1719 to test Sieve filters that use "vacation".
1721 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1722 that changes the way the GnuTLS parameters are stored in the cache file.
1723 The new format can be generated externally. For backward compatibility,
1724 if the data in the cache doesn't make sense, Exim assumes it has read an
1725 old-format file, and it generates new data and writes a new file. This
1726 means that you can't go back to an older release without removing the
1729 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1730 work if there are any delivery delays because "one_time" forces the
1731 parent to be marked "delivered", so its unseen clone is never tried
1732 again. For this reason, Exim now forbids the simultaneous setting of
1735 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1736 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1737 entry if you want to know the details. The fix, however, neglected to
1738 consider the case when local delivery batching is involved. The test for
1739 "previously delivered" was not happening when checking to see if an
1740 address could be batched with a previous (undelivered) one; under
1741 certain circumstances this could lead to multiple deliveries to the same
1744 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1745 in its include files, and this causes problems building Exim.
1747 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1748 header_syntax) but Exim was just ignoring anything given after a slash.
1749 In particular, this caused confusion with an attempt to use "verify =
1750 reverse_host_lookup/defer_ok". An error is now given when options are
1751 supplied for verify items that do not have them. (Maybe reverse_host_
1752 lookup should have a defer_ok option, but that's a different point.)
1754 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1755 defined by RFC 821) to 2048, because there were problems with some AUTH
1756 commands, and RFC 1869 says the size should be increased for extended
1757 SMTP commands that take arguments.
1759 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1762 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1763 "unknown" error; now it says that the functionality isn't in the binary.
1765 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1766 an address' error message when a string expansion fails (syntax or
1767 whatever). Otherwise the password may appear in the log. Following change
1768 PH/42 below, there is no longer a chance of it appearing in a bounce
1771 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1773 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1774 of its addresses was ever tried. (Bugzilla bug #2.)
1776 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1777 the result incorrectly in the debug output. (It correctly added a newline
1778 to what was transported.)
1780 TF/01 Added $received_time.
1782 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1783 commented out examples of how to interface to a virus scanner and to
1784 SpamAssassin. Also added commented examples of av_scanner and
1785 spamd_address settings.
1787 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1788 and controls are allowed in which ACLs. There were a couple of minor
1789 errors. Some of the entries in the conditions table (which is a table of
1790 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1791 negation of where the condition IS allowed.
1793 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1795 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1796 header file does not have a version number, so I've had to invent a new
1797 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1798 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1799 radiusclient), but it was contributed by a Radius user.
1801 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1802 files or directories, for OpenSSL.
1804 PH/19 When an Exim process that is running as root has to create an Exim log
1805 file, it does so in a subprocess that runs as exim:exim so as to get the
1806 ownership right at creation (otherwise, other Exim processes might see
1807 the file with the wrong ownership). There was no test for failure of this
1808 fork() call, which would lead to the process getting stuck as it waited
1809 for a non-existent subprocess. Forks do occasionally fail when resources
1810 run out. I reviewed all the other calls to fork(); they all seem to check
1813 PH/20 When checking for unexpected SMTP input at connect time (before writing
1814 the banner), Exim was not dealing correctly with a non-positive return
1815 from the read() function. If the client had disconnected by this time,
1816 the result was a log entry for a synchronization error with an empty
1817 string after "input=" when read() returned zero. If read() returned -1
1818 (an event I could not check), uninitialized data bytes were printed.
1819 There were reports of junk text (parts of files, etc) appearing after
1822 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1824 PH/22 Added support for macro redefinition, and (re)definition in between
1825 driver and ACL definitions.
1827 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1828 forgetting to use the resulting value; it was using the unexpanded value.
1830 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1831 hadn't been configured. The fix is from Juergen Kreileder, who
1832 understands it better than I do:
1834 "Here's what I see happening with three configured cyrus_sasl
1835 authenticators configured (plain, login, cram-md5):
1837 On startup auth_cyrus_sasl_init() gets called for each of these.
1838 This means three calls to sasl_listmech() without a specified mech_list.
1839 => SASL tests which mechs of all available mechs actually work
1840 => three warnings about OTP not working
1841 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1843 With the patch, sasl_listmech() also gets called three times. But now
1844 SASL's mech_list option is set to the server_mech specified in the the
1845 authenticator. Or in other words, the answer from sasl_listmech()
1846 gets limited to just the mech you're testing for (which is different
1848 => the return list contains just 'plain' or 'login', 'cram-md5' or
1849 nothing depending on the value of ob->server_mech.
1851 I've just tested the patch: Authentication still works fine,
1852 unavailable mechs specified in the exim configuration are still
1853 caught, and the auth.log warnings about OTP are gone."
1855 PH/25 When debugging is enabled, the contents of the command line are added
1856 to the debugging output, even when log_selector=+arguments is not
1859 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1860 answer is "GNU", and only if the return is "GNU/something" is the answer
1863 PH/27 $acl_verify_message is now set immediately after the failure of a
1864 verification in an ACL, and so is available in subsequent modifiers. In
1865 particular, the message can be preserved by coding like this:
1867 warn !verify = sender
1868 set acl_m0 = $acl_verify_message
1870 Previously, $acl_verify_message was set only while expanding "message"
1871 and "log_message" when a very denied access.
1873 PH/28 Modified OS/os.c-Linux with
1875 -#ifndef OS_LOAD_AVERAGE
1876 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1878 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1879 nomenclature these days.)
1881 PH/29 Installed patch from the Sieve maintainer that adds the options
1882 sieve_useraddress and sieve_subaddress to the redirect router.
1884 PH/30 In these circumstances:
1885 . Two addresses routed to the same list of hosts;
1886 . First host does not offer TLS;
1887 . First host accepts first address;
1888 . First host gives temporary error to second address;
1889 . Second host offers TLS and a TLS session is established;
1890 . Second host accepts second address.
1891 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1892 and peerdn, if requested) that were in fact used only for the second
1895 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1896 attention to any local part prefix or suffix that was matched by the
1897 router that accepted the address. It now behaves in the same way as it
1898 does for delivery: the affixes are removed from the local part unless
1899 rcpt_include_affixes is set on the transport.
1901 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1902 timeout during the DATA phase of an incoming message.
1904 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1905 applied a patch sanctioned by the Sieve maintainer.
1907 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1908 the uid or gid is negative. A case of a negative gid caused this to be
1909 noticed. The fix allows for either to be negative.
1911 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1912 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1915 PH/36 The change PH/12 above was broken. Fixed it.
1917 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1918 the grounds that routing the same address twice would always produce the
1919 same answer. This might have been true once, but it is certainly no
1920 longer true now. Routing a child address may depend on the previous
1921 routing that produced that child. Some complicated redirection strategies
1922 went wrong when messages had multiple recipients, and made Exim's
1923 behaviour dependent on the order in which the addresses were given.
1925 I have moved the duplicate checking until after the routing is complete.
1926 Exim scans the addresses that are assigned to local and remote
1927 transports, and removes any duplicates. This means that more work will be
1928 done, as duplicates will always all be routed, but duplicates are
1929 presumably rare, so I don't expect this is of any significance.
1931 For deliveries to pipes, files, and autoreplies, the duplicate checking
1932 still happens during the routing process, since they are not going to be
1935 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1936 It corrects a timeout issue with spamd. This is Ian's comment: "The
1937 background is that sometimes spamd either never reads data from a
1938 connection it has accepted, or it never writes response data. The exiscan
1939 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1940 blindly assumes that writes won't block so it may never time out."
1942 PH/39 Allow G after quota size as well as K and M.
1944 PH/40 The value set for $authenticated_id in an authenticator may not contain
1945 binary zeroes or newlines because the value is written to log lines and
1946 to spool files. There was no check on this. Now the value is run through
1947 the string_printing() function so that such characters are converted to
1948 printable escape sequences.
1950 PH/41 $message_linecount is a new variable that contains the total number of
1951 lines in the message. Compare $body_linecount, which is the count for the
1954 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1955 bounce and delay warning messages, except in certain special cases, which
1958 (a) An SMTP error message from a remote host;
1959 (b) A message specified in a :fail: redirection;
1960 (c) A message specified in a "fail" command in a system filter;
1961 (d) A message specified in a FAIL return from the queryprogram router;
1962 (e) A message specified by the cannot_route_message router option.
1964 In these cases only, Exim does include the error details in bounce and
1965 warning messages. There are also a few cases where bland messages such
1966 as "unrouteable address" or "local delivery error" are given.
1968 PH/43 $value is now also set for the "else" part of a ${run expansion.
1970 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1971 being worked on, but at least Exim now implements the latest version to
1974 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1975 process to complete was treated as a delivery failure, a timeout while
1976 writing the message to the pipe was logged, but erroneously treated as a
1977 successful delivery. Such timeouts include transport filter timeouts. For
1978 consistency with the overall process timeout, these timeouts are now
1979 treated as errors, giving rise to delivery failures by default. However,
1980 there is now a new Boolean option for the pipe transport called
1981 timeout_defer, which, if set TRUE, converts the failures into defers for
1982 both kinds of timeout. A transport filter timeout is now identified in
1985 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1986 systems where "make" and "gmake" are different, calling "gmake" at top
1987 level broke things. I've arranged for the value of $(MAKE) to be passed
1988 from the Makefile to this script so that it can call the same version of
1992 A note about Exim versions 4.44 and 4.50
1993 ----------------------------------------
1995 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1996 changes of various kinds. As a consequence, a big documentation update was
1997 needed. This delayed the release for rather longer than seemed good, especially
1998 in the light of a couple of (minor) security issues. Therefore, the changes
1999 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2000 release. So 4.44 and 4.50 are in effect two different branches that both start
2003 I have left the 4.50 change log unchanged; it contains all the changes since
2004 4.43. The change log for 4.44 is below; many of its items are identical to
2005 those for 4.50. This seems to be the most sensible way to preserve the
2006 historical information.
2012 1. Minor wording change to the doc/README.SIEVE file.
2014 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2015 computation of the current number of files was incorrect.
2017 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2018 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2021 4. Give more explanation in the error message when the command for a transport
2022 filter fails to execute.
2024 5. There are several places where Exim runs a non-Exim command in a
2025 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2026 was being done only for the command run by the queryprogram router. It is
2027 now done for all such subprocesses. The other cases are: ${run, transport
2028 filters, and the commands run by the lmtp and pipe transports.
2030 6. Added CONFIGURE_GROUP build-time option.
2032 7. Some older OS have a limit of 256 on the maximum number of file
2033 descriptors. Exim was using setrlimit() to set 1000 as a large value
2034 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2035 systems. I've change it so that if it can't get 1000, it tries for 256.
2037 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2038 was an oversight, and furthermore, ever since the addition of extra
2039 controls (e.g. 4.43/32), the checks on when to allow different forms of
2040 "control" were broken. There should now be diagnostics for all cases when a
2041 control that does not make sense is encountered.
2043 9. Added the /retain_sender option to "control=submission".
2045 10. $recipients is now available in the predata ACL (oversight).
2047 11. Tidy the search cache before the fork to do a delivery from a message
2048 received from the command line. Otherwise the child will trigger a lookup
2049 failure and thereby defer the delivery if it tries to use (for example) a
2050 cached ldap connection that the parent has called unbind on.
2052 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2053 of $address_data from the recipient verification was clobbered by the
2054 sender verification.
2056 13. The value of address_data from a sender verification is now available in
2057 $sender_address_data in subsequent conditions in the ACL statement.
2059 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2061 15. Added a new option "connect=<time>" to callout options, to set a different
2064 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2065 was its contents. (It was OK if the option was not defined at all.)
2067 17. A "Completed" log line is now written for messages that are removed from
2068 the spool by the -Mrm option.
2070 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2071 information about exactly what failed.
2073 19. Added -dd to debug only the daemon process.
2075 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2076 handles timeouts, both on the server side and network timeouts. Renamed the
2077 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2079 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2080 It is now set to "smtps".
2082 22. $host_address is now set to the target address during the checking of
2083 ignore_target_hosts.
2085 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2086 being passed; this would have caused $sender_host_name to have been used if
2087 matching the list had actually called for a host name (not very likely,
2088 since this list is usually IP addresses). A host name is now passed as
2091 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2092 code that specifies a non-restarting handler (typically sigaction() in
2093 modern systems) in an attempt to fix a rare and obscure crash bug.
2095 25. Narrowed the window for a race in the daemon that could cause it to ignore
2096 SIGCHLD signals. This is not a major problem, because they are used only to
2097 wake it up if nothing else does.
2099 26. A malformed maildirsize file could cause Exim to calculate negative values
2100 for the mailbox size or file count. Odd effects could occur as a result.
2101 The maildirsize information is now recalculated if the size or filecount
2104 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2105 support for a long time. Removed HAVE_SYS_VFS_H.
2107 28. Installed the latest version of exipick from John Jetmore.
2109 29. In an address list, if the pattern was not a regular expression, an empty
2110 subject address (from a bounce message) matched only if the pattern was an
2111 empty string. Non-empty patterns were not even tested. This was the wrong
2112 because it is perfectly reasonable to use an empty address as part of a
2113 database query. An empty address is now tested by patterns that are
2114 lookups. However, all the other forms of pattern expect the subject to
2115 contain a local part and a domain, and therefore, for them, an empty
2116 address still always fails if the pattern is not itself empty.
2118 30. Exim went into a mad DNS loop when attempting to do a callout where the
2119 host was specified on an smtp transport, and looking it up yielded more
2120 than one IP address.
2122 31. Re-factored the code for checking spool and log partition space into a
2123 function that finds that data and another that does the check. The former
2124 is then used to implement four new variables: $spool_space, $log_space,
2125 $spool_inodes, and $log_inodes.
2127 32. The RFC2047 encoding function was originally intended for short strings
2128 such as real names; it was not keeping to the 75-character limit for
2129 encoded words that the RFC imposes. It now respects the limit, and
2130 generates multiple encoded words if necessary. To be on the safe side, I
2131 have increased the buffer size for the ${rfc2047: expansion operator from
2134 33. It is now permitted to omit both strings after an "if" condition; if the
2135 condition is true, the result is "true". As before, when the second string
2136 is omitted, a false condition yields an empty string. This makes it less
2137 cumbersome to write custom ACL and router conditions.
2139 34. Failure to deliver a bounce message always caused it to be frozen, even if
2140 there was an errors_to setting on the router. The errors_to setting is now
2143 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2144 canonical form (fully expanded) before being placed in
2145 $sender_host_address.
2147 36. The table in the code that translates DNS record types into text (T_A to
2148 "A" for instance) was missing entries for NS and CNAME. It is just possible
2149 that this could have caused confusion if both these types were looked up
2150 for the same domain, because the text type is used as part of Exim's
2151 per-process caching. But the chance of anyone hitting this buglet seems
2154 37. The dnsdb lookup has been extended in a number of ways.
2156 (1) There is a new type, "zns", which walks up the domain tree until it
2157 finds some nameserver records. It should be used with care.
2159 (2) There is a new type, "mxh", which is like "mx" except that it returns
2160 just the host names, not the priorities.
2162 (3) It is now possible to give a list of domains (or IP addresses) to be
2163 looked up. The behaviour when one of the lookups defers can be
2164 controlled by a keyword.
2166 (4) It is now possible to specify the separator character for use when
2167 multiple records are returned.
2169 38. The dnslists ACL condition has been extended: it is now possible to supply
2170 a list of IP addresses and/or domains to be looked up in a particular DNS
2173 39. Added log_selector=+queue_time_overall.
2175 40. When running the queue in the test harness, wait just a tad after forking a
2176 delivery process, to get repeatability of debugging output.
2178 41. Include certificate and key file names in error message when GnuTLS fails
2179 to set them up, because the GnuTLS error message doesn't include the name
2180 of the failing file when there is a problem reading it.
2182 42. Allow both -bf and -bF in the same test run.
2184 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2186 44. The "Exiscan patch" is now merged into the mainline Exim source.
2188 45. Sometimes the final signoff response after QUIT could fail to get
2189 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2190 < 0 before doing a fflush(). This bug looks as though it goes back to the
2191 introduction of TLS in release 3.20, but "sometimes" must have been rare
2192 because the tests only now provoked it.
2194 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2195 (this can affect the format of dates).
2197 47. exim_tidydb, when checking for the continued existence of a message for
2198 which it has found a message-specific retry record, was not finding
2199 messages that were in split spool directories. Consequently, it was
2200 deleting retry records that should have stayed in existence.
2202 48. Steve fixed some bugs in eximstats.
2204 49. The SPA authentication driver was not abandoning authentication and moving
2205 on to the next authenticator when an expansion was forced to fail,
2206 contradicting the general specification for all authenticators. Instead it
2207 was generating a temporary error. It now behaves as specified.
2209 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2210 (the order specifies the preference for clients). The order is now AES256,
2211 AES128, 3DES, ARCFOUR128.
2213 51. Small patch to Sieve code - explicitly set From: when generating an
2216 52. Exim crashed if a remote delivery caused a very long error message to be
2217 recorded - for instance if somebody sent an entire SpamAssassin report back
2218 as a large number of 550 error lines. This bug was coincidentally fixed by
2219 increasing the size of one of Exim's internal buffers (big_buffer) that
2220 happened as part of the Exiscan merge. However, to be on the safe side, I
2221 have made the code more robust (and fixed the comments that describe what
2224 53. Now that there can be additional text after "Completed" in log lines (if
2225 the queue_time_overall log selector is set), a one-byte patch to exigrep
2226 was needed to allow it to recognize "Completed" as not the last thing in
2229 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2230 patch that reportedly fixes this has been added. I am not expert enough to
2231 create a test for it. This is what the patch creator wrote:
2233 "I found a little strange behaviour of ldap code when working with
2234 Windows 2003 AD Domain, where users was placed in more than one
2235 Organization Units. When I tried to give exim partial DN, the exit code
2236 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2237 But simultaneously result of request was absolutely normal ldap result,
2238 so I produce this patch..."
2240 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2241 REFERENCE, so I have modified the code to exclude the patch when that macro
2244 55. Some experimental protocols are using DNS PTR records for new purposes. The
2245 keys for these records are domain names, not reversed IP addresses. The
2246 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2247 leaves it alone. Component reversal etc. now happens only for IP addresses.
2250 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2252 57. Double the size of the debug message buffer (to 2048) so that more of very
2253 long debug lines gets shown.
2255 58. The exicyclog utility now does better if the number of log files to keep
2256 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2258 59. Two changes related to the smtp_active_hostname option:
2260 (1) $smtp_active_hostname is now available as a variable.
2261 (2) The default for smtp_banner uses $smtp_active_hostname instead
2262 of $primary_hostname.
2264 60. The host_aton() function is supposed to be passed a string that is known
2265 to be a valid IP address. However, in the case of IPv6 addresses, it was
2266 not checking this. This is a hostage to fortune. Exim now panics and dies
2267 if the condition is not met. A case was found where this could be provoked
2268 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2269 components; fortuitously, this particular loophole had already been fixed
2270 by change 4.50/55 above.
2272 If there are any other similar loopholes, the new check in host_aton()
2273 itself should stop them being exploited. The report I received stated that
2274 data on the command line could provoke the exploit when Exim was running as
2275 exim, but did not say which command line option was involved. All I could
2276 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2277 running as the user.
2280 61. There was a buffer overflow vulnerability in the SPA authentication code
2281 (which came originally from the Samba project). I have added a test to the
2282 spa_base64_to_bits() function which I hope fixes it.
2285 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2286 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2288 63. The daemon start-up calls getloadavg() while still root for those OS that
2289 need the first call to be done as root, but it missed one case: when
2290 deliver_queue_load_max is set with deliver_drop_privilege. This is
2291 necessary for the benefit of the queue runner, because there is no re-exec
2292 when deliver_drop_privilege is set.
2294 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2295 This has been fixed.
2297 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2298 was in use, was not putting the data itself into the right store pool;
2299 consequently, it could be overwritten for a subsequent message in the same
2300 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2303 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2305 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2306 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2307 the function were treating the return as a boolean value, which happened to
2308 work because 0=false and not-0=true, but is not correct code.
2310 68. The host_aton() function was not handling scoped IPv6 addresses (those
2311 with, for example, "%eth0" on the end) correctly.
2313 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2314 negated items (that is, ~something) in unsigned ints. Some compilers
2315 apparently mutter when there is no cast.
2317 70. If an address verification called from an ACL failed, and did not produce a
2318 user-specific message (i.e. there was only a "system" message), nothing was
2319 put in $acl_verify_message. In this situation, it now puts the system
2322 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2323 session; change 4.31/43 added the unwanted input to the log line - except
2324 that it did not do this in the start of session case. It now does.
2326 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2327 This is wrong and can cause the other end to generate a synchronization
2328 error if it is another Exim or anything else that does the synchronization
2329 check. A QUIT command is no longer sent after a timeout.
2331 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2332 during host lookups.
2334 74. The defer_ok option of callout verification was not working if it was used
2335 when verifying addresses in header lines, that is, for this case:
2337 verify = header_sender/callout=defer_ok
2339 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2340 those file descriptors could be used for SMTP connections. If anything
2341 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2342 could be sent to the SMTP client, causing chaos. The daemon now opens
2343 stdin, stdout, and stderr to /dev/null when it puts itself into the
2346 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2347 log by default. The user can override this with suitable Perl magic.
2349 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2350 the log message when discard triggers, was not working for the DATA ACL or
2351 for the non-SMTP ACL.
2353 78. Error message wording change in sieve.c.
2355 79. If smtp_accept_max_per_host was set, the number of connections could be
2356 restricted to fewer than expected, because the daemon was trying to set up
2357 a new connection before checking whether the processes handling previous
2358 connections had finished. The check for completed processes is now done
2359 earlier. On busy systems, this bug wouldn't be noticed because something
2360 else would have woken the daemon, and it would have reaped the completed
2363 80. If a message was submitted locally by a user whose login name contained one
2364 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2365 It caused a spool format error. I have fixed the spool reading code. A
2366 related problem was that the "from" clause in the Received: line became
2367 illegal because of the space(s). It is now covered by ${quote_local_part.
2369 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2372 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2374 83. Patch from Sieve maintainer to fix unterminated string problem in
2375 "vacation" handling.
2377 84. Some minor changes to the Linux configuration files to help with other
2378 OS variants using glibc.
2380 85. One more patch for Sieve to update vacation handling to latest spec.
2383 ----------------------------------------------------
2384 See the note above about the 4.44 and 4.50 releases.
2385 ----------------------------------------------------
2391 1. Change 4.43/35 introduced a bug that caused file counts to be
2392 incorrectly computed when quota_filecount was set in an appendfile
2395 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2396 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2399 3. Give more explanation in the error message when the command for a transport
2400 filter fails to execute.
2402 4. There are several places where Exim runs a non-Exim command in a
2403 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2404 was being done only for the command run by the queryprogram router. It is
2405 now done for all such subprocesses. The other cases are: ${run, transport
2406 filters, and the commands run by the lmtp and pipe transports.
2408 5. Some older OS have a limit of 256 on the maximum number of file
2409 descriptors. Exim was using setrlimit() to set 1000 as a large value
2410 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2411 systems. I've change it so that if it can't get 1000, it tries for 256.
2413 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2414 was an oversight, and furthermore, ever since the addition of extra
2415 controls (e.g. 4.43/32), the checks on when to allow different forms of
2416 "control" were broken. There should now be diagnostics for all cases when a
2417 control that does not make sense is encountered.
2419 7. $recipients is now available in the predata ACL (oversight).
2421 8. Tidy the search cache before the fork to do a delivery from a message
2422 received from the command line. Otherwise the child will trigger a lookup
2423 failure and thereby defer the delivery if it tries to use (for example) a
2424 cached ldap connection that the parent has called unbind on.
2426 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2427 of $address_data from the recipient verification was clobbered by the
2428 sender verification.
2430 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2431 was its contents. (It was OK if the option was not defined at all.)
2433 11. A "Completed" log line is now written for messages that are removed from
2434 the spool by the -Mrm option.
2436 12. $host_address is now set to the target address during the checking of
2437 ignore_target_hosts.
2439 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2440 being passed; this would have caused $sender_host_name to have been used if
2441 matching the list had actually called for a host name (not very likely,
2442 since this list is usually IP addresses). A host name is now passed as
2445 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2446 code that specifies a non-restarting handler (typically sigaction() in
2447 modern systems) in an attempt to fix a rare and obscure crash bug.
2449 15. Narrowed the window for a race in the daemon that could cause it to ignore
2450 SIGCHLD signals. This is not a major problem, because they are used only to
2451 wake it up if nothing else does.
2453 16. A malformed maildirsize file could cause Exim to calculate negative values
2454 for the mailbox size or file count. Odd effects could occur as a result.
2455 The maildirsize information is now recalculated if the size or filecount
2458 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2459 support for a long time. Removed HAVE_SYS_VFS_H.
2461 18. Updated exipick to current release from John Jetmore.
2463 19. Allow an empty sender to be matched against a lookup in an address list.
2464 Previously the only cases considered were a regular expression, or an
2467 20. Exim went into a mad DNS lookup loop when doing a callout where the
2468 host was specified on the transport, if the DNS lookup yielded more than
2471 21. The RFC2047 encoding function was originally intended for short strings
2472 such as real names; it was not keeping to the 75-character limit for
2473 encoded words that the RFC imposes. It now respects the limit, and
2474 generates multiple encoded words if necessary. To be on the safe side, I
2475 have increased the buffer size for the ${rfc2047: expansion operator from
2478 22. Failure to deliver a bounce message always caused it to be frozen, even if
2479 there was an errors_to setting on the router. The errors_to setting is now
2482 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2483 canonical form (fully expanded) before being placed in
2484 $sender_host_address.
2486 24. Updated eximstats to version 1.33
2488 25. Include certificate and key file names in error message when GnuTLS fails
2489 to set them up, because the GnuTLS error message doesn't include the name
2490 of the failing file when there is a problem reading it.
2492 26. Expand error message when OpenSSL has problems setting up cert/key files.
2495 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2496 (this can affect the format of dates).
2498 28. exim_tidydb, when checking for the continued existence of a message for
2499 which it has found a message-specific retry record, was not finding
2500 messages that were in split spool directories. Consequently, it was
2501 deleting retry records that should have stayed in existence.
2503 29. eximstats updated to version 1.35
2504 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2505 1.35 - bugfix such that pie charts by volume are generated correctly
2507 30. The SPA authentication driver was not abandoning authentication and moving
2508 on to the next authenticator when an expansion was forced to fail,
2509 contradicting the general specification for all authenticators. Instead it
2510 was generating a temporary error. It now behaves as specified.
2512 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2513 (the order specifies the preference for clients). The order is now AES256,
2514 AES128, 3DES, ARCFOUR128.
2516 31. Small patch to Sieve code - explicitly set From: when generating an
2519 32. Exim crashed if a remote delivery caused a very long error message to be
2520 recorded - for instance if somebody sent an entire SpamAssassin report back
2521 as a large number of 550 error lines. This bug was coincidentally fixed by
2522 increasing the size of one of Exim's internal buffers (big_buffer) that
2523 happened as part of the Exiscan merge. However, to be on the safe side, I
2524 have made the code more robust (and fixed the comments that describe what
2527 33. Some experimental protocols are using DNS PTR records for new purposes. The
2528 keys for these records are domain names, not reversed IP addresses. The
2529 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2530 leaves it alone. Component reversal etc. now happens only for IP addresses.
2533 34. The host_aton() function is supposed to be passed a string that is known
2534 to be a valid IP address. However, in the case of IPv6 addresses, it was
2535 not checking this. This is a hostage to fortune. Exim now panics and dies
2536 if the condition is not met. A case was found where this could be provoked
2537 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2538 components; fortuitously, this particular loophole had already been fixed
2539 by change 4.50/55 or 4.44/33 above.
2541 If there are any other similar loopholes, the new check in host_aton()
2542 itself should stop them being exploited. The report I received stated that
2543 data on the command line could provoke the exploit when Exim was running as
2544 exim, but did not say which command line option was involved. All I could
2545 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2546 running as the user.
2549 35. There was a buffer overflow vulnerability in the SPA authentication code
2550 (which came originally from the Samba project). I have added a test to the
2551 spa_base64_to_bits() function which I hope fixes it.
2554 36. The daemon start-up calls getloadavg() while still root for those OS that
2555 need the first call to be done as root, but it missed one case: when
2556 deliver_queue_load_max is set with deliver_drop_privilege. This is
2557 necessary for the benefit of the queue runner, because there is no re-exec
2558 when deliver_drop_privilege is set.
2560 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2561 was in use, was not putting the data itself into the right store pool;
2562 consequently, it could be overwritten for a subsequent message in the same
2563 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2566 38. Sometimes the final signoff response after QUIT could fail to get
2567 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2568 < 0 before doing a fflush(). This bug looks as though it goes back to the
2569 introduction of TLS in release 3.20, but "sometimes" must have been rare
2570 because the tests only now provoked it.
2576 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2577 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2578 Now it returns an integer. A number of calls were still expecting a T/F
2579 return. Fortuitously, in all cases, the tests worked in OK situations,
2580 which is the norm. However, things would have gone wrong on any write
2581 failures on the smtp file descriptor. This function is used when sending
2582 messages over SMTP and also when doing verify callouts.
2584 2. When Exim is called to do synchronous delivery of a locally submitted
2585 message (the -odf or -odi options), it no longer closes stderr before doing
2588 3. Implemented the mua_wrapper option.
2590 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2592 5. Implemented the functions header_remove(), header_testname(),
2593 header_add_at_position(), and receive_remove_recipient(), and exported them
2596 6. If an ACL "warn" statement specified the addition of headers, Exim already
2597 inserted X-ACL-Warn: at the start if there was no header name. However, it
2598 was not making this test for the second and subsequent header lines if
2599 there were newlines in the string. This meant that an invalid header could
2600 be inserted if Exim was badly configured.
2602 7. Allow an ACL "warn" statement to add header lines at the start or after all
2603 the Received: headers, as well as at the end.
2605 8. Added the rcpt_4xx retry error code.
2607 9. Added postmaster_mailfrom=xxx to callout verification option.
2609 10. Added mailfrom=xxxx to the callout verification option, for verify=
2612 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2613 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2615 12. Inserted some casts to stop certain compilers warning when using pointer
2616 differences as field lengths or precisions in printf-type calls (mostly
2617 affecting debugging statements).
2619 13. Added optional readline() support for -be (dynamically loaded).
2621 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2622 same clock tick as a message's arrival, so that its received time was the
2623 same as the "first fail" time on the retry record, and that message
2624 remained on the queue past the ultimate address timeout, every queue runner
2625 would try a delivery (because it was past the ultimate address timeout) but
2626 after another failure, the ultimate address timeout, which should have then
2627 bounced the address, did not kick in. This was a "< instead of <=" error;
2628 in most cases the first failure would have been in the next clock tick
2629 after the received time, and all would be well.
2631 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2632 being recognized when the domain list was tested by the match_domain
2633 condition in an expansion string.
2635 16. Added the ${str2b64: operator.
2637 17. Exim was always calling setrlimit() to set a large limit for the number of
2638 processes, without checking whether the existing limit was already
2639 adequate. (It did check for the limit on file descriptors.) Furthermore,
2640 errors from getrlimit() and setrlimit() were being ignored. Now they are
2641 logged to the main and panic logs, but Exim does carry on, to try to do its
2642 job under whatever limits there are.
2644 18. Imported PCRE 5.0.
2646 19. Trivial typo in log message " temporarily refused connection" (the leading
2649 20. If the log selector return_path_on_delivery was set and an address was
2650 redirected to /dev/null, the delivery process crashed because it assumed
2651 that a return path would always be set for a "successful" delivery. In this
2652 case, the whole delivery is bypassed as an optimization, and therefore no
2655 21. Internal re-arrangement: the function for sending a challenge and reading
2656 a response while authentication was assuming a zero-terminated challenge
2657 string. It's now changed to take a pointer and a length, to allow for
2658 binary data in such strings.
2660 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2662 23. Exim was not respecting finduser_retries when seeking the login of the
2663 uid under which it was called; it was always trying 10 times. (The default
2664 setting of finduser_retries is zero.) Also, it was sleeping after the final
2665 failure, which is pointless.
2667 24. Implemented tls_on_connect_ports.
2669 25. Implemented acl_smtp_predata.
2671 26. If the domain in control=submission is set empty, Exim assumes that the
2672 authenticated id is a complete email address when it generates From: or
2673 Sender: header lines.
2675 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2676 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2677 chown and chgrp in /bin and hostname in /usr/bin.
2679 28. Exim was keeping the "process log" file open after each use, just as it
2680 does for the main log. This opens the possibility of it remaining open for
2681 long periods when the USR1 signal hits a daemon. Occasional processlog
2682 errors were reported, that could have been caused by this. Anyway, it seems
2683 much more sensible not to leave this file open at all, so that is what now
2686 29. The long-running daemon process does not normally write to the log once it
2687 has entered its main loop, and it closes the log before doing so. This is
2688 so that log files can straightforwardly be renamed and moved. However,
2689 there are a couple of unusual error situations where the daemon does write
2690 log entries, and I had neglected to close the log afterwards.
2692 30. The text of an SMTP error response that was received during a remote
2693 delivery was being truncated at 512 bytes. This is too short for some of
2694 the long messages that one sometimes sees. I've increased the limit to
2697 31. It is now possible to make retry rules that apply only when a message has a
2698 specific sender, in particular, an empty sender.
2700 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2701 it possible to be selective about when SMTP synchronization is enforced.
2703 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2705 32. Implemented hosts_connection_nolog.
2707 33. Added an ACL for QUIT.
2709 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2712 35. Added mailbox_size and mailbox_filecount to appendfile.
2714 36. Added control = no_multiline_responses to ACLs.
2716 37. There was a bug in the logic of the code that waits for the clock to tick
2717 in the case where the clock went backwards by a substantial amount such
2718 that the microsecond fraction of "now" was more than the microsecond
2719 fraction of "then" (but the whole seconds number was less).
2721 38. Added support for the libradius Radius client library this is found on
2722 FreeBSD (previously only the radiusclient library was supported).
2728 1. When certain lookups returned multiple values in the form name=value, the
2729 quoting of the values was not always being done properly. Specifically:
2730 (a) If the value started with a double quote, but contained no whitespace,
2732 (b) If the value contained whitespace other than a space character (i.e.
2733 tabs or newlines or carriage returns) it was not quoted.
2734 This fix has been applied to the mysql and pgsql lookups by writing a
2735 separate quoting function and calling it from the lookup code. The fix
2736 should probably also be applied to nisplus, ibase and oracle lookups, but
2737 since I cannot test any of those, I have not disturbed their existing code.
2739 2. A hit in the callout cache for a specific address caused a log line with no
2740 reason for rejecting RCPT. Now it says "Previous (cached) callout
2741 verification failure".
2743 3. There was an off-by-one bug in the queryprogram router. An over-long
2744 return line was truncated at 256 instead of 255 characters, thereby
2745 overflowing its buffer with the terminating zero. As well as fixing this, I
2746 have increased the buffer size to 1024 (and made a note to document this).
2748 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2749 when Exim is waiting for an SMTP response from a remote server, Exim
2750 restarts its select() call on the socket, thereby resetting its timeout.
2751 This is not a problem when such interrupts are rare. Somebody set up a cron
2752 job to run exiwhat every 2 minutes, which is less than the normal select()
2753 timeout (5 or 10 minutes). This meant that the select() timeout never
2754 kicked in because it was always reset. I have fixed this by comparing the
2755 time when an interrupt arrives with the time at the start of the first call
2756 to select(). If more time than the timeout has elapsed, the interrupt is
2757 treated as a timeout.
2759 5. Some internal re-factoring in preparation for the addition of Sieve
2760 extensions (by MH). In particular, the "personal" test is moved to a
2761 separate function, and given an option for scanning Cc: and Bcc: (which is
2762 not set for Exim filters).
2764 6. When Exim created an email address using the login of the caller as the
2765 local part (e.g. when creating a From: or Sender: header line), it was not
2766 quoting the local part when it contained special characters such as @.
2768 7. Installed new OpenBSD configuration files.
2770 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2771 try to make them clearer.
2773 9. Callout options, other than the timeout value, were being ignored when
2774 verifying sender addresses in header lines. For example, when using
2776 verify = header_sender/callout=no_cache
2778 the cache was (incorrectly) being used.
2780 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2781 only the Cygwin environment.
2783 11. When return_path_on_delivery was set as a log selector, if different remote
2784 addresses in the same message used different return paths and parallel
2785 remote delivery occurred, the wrong values would sometimes be logged.
2786 (Whenever a remote delivery process finished, the return path value from
2787 the most recently started remote delivery process was logged.)
2789 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2790 lines when AUTH and/or TLS are in use. This is the "received protocol"
2791 field. Exim used to use "asmtp" for authenticated SMTP, without any
2792 indication (in the protocol name) for TLS use. Now it follows the RFC and
2793 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2794 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2795 names appear in log lines as well as in Received: header lines.
2797 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2798 extensions, and comparison tests, and to fix some bugs.
2800 14. Changes to the "personal" filter test:
2802 (1) The test was buggy in that it was just doing the equivalent of
2803 "contains" tests on header lines. For example, if a user's address was
2804 anne@some.where, the "personal" test would incorrectly be true for
2806 To: susanne@some.where
2808 This test is now done by extracting each address from the header in turn,
2809 and checking the entire address. Other tests that are part of "personal"
2810 are now done using regular expressions (for example, to check local parts
2811 of addresses in From: header lines).
2813 (2) The list of non-personal local parts in From: addresses has been
2814 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2815 taken from the Sieve specification recommendations.
2817 (3) If the message contains any header line starting with "List-" it is
2818 treated as non-personal.
2820 (4) The test for "circular" in the Subject: header line has been removed
2821 because it now seems ill-conceived.
2823 15. Minor typos in src/EDITME comments corrected.
2825 16. Installed latest exipick from John Jetmore.
2827 17. If headers_add on a router specified a text string that was too long for
2828 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2829 of string_sprintf() is now avoided.
2831 18. $message_body_size was not set (it was always zero) when running the DATA
2832 ACL and the local_scan() function.
2834 19. For the "mail" command in an Exim filter, no default was being set for
2835 the once_repeat time, causing a random time value to be used if "once" was
2836 specified. (If the value happened to be <= 0, no repeat happened.) The
2837 default is now 0s, meaning "never repeat". The "vacation" command was OK
2838 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2839 (I found it when inspecting the code).
2841 20. There is now an overall timeout for performing a callout verification. It
2842 defaults to 4 times the callout timeout, which applies to individual SMTP
2843 commands during the callout. The overall timeout applies when there is more
2844 than one host that can be tried. The timeout is checked before trying the
2845 next host. This prevents very long delays if there are a large number of
2846 hosts and all are timing out (e.g. when the network connections are timing
2847 out). The value of the overall timeout can be changed by specifying an
2848 additional sub-option for "callout", called "maxwait". For example:
2850 verify = sender/callout=5s,maxwait=20s
2852 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2853 to the end before writing, but this should make it even safer).
2855 22. Exim was forgetting that it had advertised PIPELINING for the second and
2856 subsequent messages on an SMTP connection. It was also not resetting its
2857 memory on STARTTLS and an internal HELO.
2859 23. When Exim logs an SMTP synchronization error within a session, it now
2860 records whether PIPELINING has been advertised or not.
2862 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2863 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2864 rather than long int.
2866 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2868 26. Added the never_mail option to autoreply.
2874 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2875 crash if the getsockname() call failed; this can happen if a connection is
2876 closed very soon after it is established. The problem was simply in the
2877 order in which certain operations were done, causing Exim to try to write
2878 to the SMTP stream before it had set up the file descriptor. The bug has
2879 been fixed by making things happen in the correct order.
2885 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2886 before the connection was closed, thus losing the rejection response.
2888 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2889 some early Solaris releases, but causes trouble in current releases where
2890 socklen_t is defined.
2892 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2895 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2898 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2899 that releasing the top of it at the end releases what was used for sub-
2900 expansions (unless the block got too big). However, discard this block if
2901 the first thing is a variable or header, so that we can use its block when
2902 it is dynamic (useful for very large $message_headers, for example).
2904 6. Lookups now cache *every* query, not just the most recent. A new, separate
2905 store pool is used for this. It can be recovered when all lookup caches are
2906 flushed. Lookups now release memory at the end of their result strings.
2907 This has involved some general refactoring of the lookup sources.
2909 7. Some code has been added to the store_xxx() functions to reduce the amount
2910 of flapping under certain conditions.
2912 8. log_incoming_interface used to affect only the <= reception log lines. Now
2913 it causes the local interface and port to be added to several more SMTP log
2914 lines, for example "SMTP connection from", and rejection lines.
2916 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2918 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2920 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2921 could be overwritten at the end of the current message (or the start of a
2922 new message if it was set in a HELO ACL). The value is now preserved for
2923 the duration of the SMTP connection.
2925 12. If a transport had a headers_rewrite setting, and a matching header line
2926 contained an unqualified address, that address was qualified, even if it
2927 did not match any rewriting rules. The underlying bug was that the values
2928 of the flags that permit the existence of unqualified sender and recipient
2929 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2930 non-local messages, and by -bnq for local messages) were not being
2931 preserved with the message after it was received.
2933 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2934 "next input=" as part of the text comprising the host identity instead of
2935 the correct text. The code was using the same buffer for two different
2936 strings. However, depending on which order the printing function evaluated
2937 its arguments, the bug did not always show up. Under Linux, for example, my
2938 test suite worked just fine.
2940 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2941 doesn't work with some older versions of Perl. It has been changed to "my",
2942 which in any case is probably the better facility to use.
2944 15. A really picky compiler found some instances of statements for creating
2945 error messages that either had too many or two few arguments for the format
2948 16. The size of the buffer for calls to the DNS resolver has been increased
2949 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2950 for addresses that have a lot of PTR records. This alleviates a problem; it
2951 does not fully solve it.
2953 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2954 buffer now truncates the list and logs the incident, which is the same
2955 action as happens when Exim is looking up a host name and its aliases.
2956 Previously in this situation something unpredictable would happen;
2957 sometimes it was "internal error: store_reset failed".
2959 18. If a server dropped the connection unexpectedly when an Exim client was
2960 using GnuTLS and trying to read a response, the client delivery process
2961 crashed while trying to generate an error log message.
2963 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2964 string, for example:
2966 warn message = H1: something\nH2: something
2968 the text was added as a single header line from Exim's point of view
2969 though it ended up OK in the delivered message. However, searching for the
2970 second and subsequent header lines using $h_h2: did not work. This has been
2971 fixed. Similarly, if a system filter added multiple headers in this way,
2972 the routers could not see them.
2974 20. Expanded the error message when iplsearch is called with an invalid key to
2975 suggest using net-iplsearch in a host list.
2977 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2978 ACLs are no longer actually imposed (and a message to that effect is
2981 22. If a "gecos" field in a passwd entry contained escaped characters, in
2982 particular, if it contained a \" sequence, Exim got it wrong when building
2983 a From: or a Sender: header from that name. A second bug also caused
2984 incorrect handling when an unquoted " was present following a character
2985 that needed quoting.
2987 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2988 was not being matched caselessly.
2990 24. Arranged for all hyphens in the exim.8 source to be escaped with
2993 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2994 properly. Recipient callout cache records were still being keyed to include
2995 the sender, even when use_sender was set false. This led to far more
2996 callouts that were necessary. The sender is no longer included in the key
2997 when use_sender is false.
2999 26. Added "control = submission" modifier to ACLs.
3001 27. Added the ${base62d: operator to decode base 62 numbers.
3003 28. dnsdb lookups can now access SRV records.
3005 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3006 the configuration file.
3008 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3009 (-v) mode. This makes the output for a verbose queue run more intelligible.
3011 31. Added a use_postmaster feature to recipient callouts.
3013 32. Added the $body_zerocount variable, containing the number of binary zero
3014 bytes in the message body.
3016 33. The time of last modification of the "new" subdirectory is now used as the
3017 "mailbox time last read" when there is a quota error for a maildir
3020 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3022 35. Added +ignore_unknown as a special item in host lists.
3024 36. Code for decoding IPv6 addresses in host lists is now included, even if
3025 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3026 address was recognized as an IP address, but was then not correctly decoded
3027 into binary, causing unexpected and incorrect effects when compared with
3034 1. Very minor rewording of debugging text in manualroute to say "list of
3035 hosts" instead of "hostlist".
3037 2. If verify=header_syntax was set, and a header line with an unqualified
3038 address (no domain) and a large number of spaces between the end of the
3039 name and the colon was received, the reception process suffered a buffer
3040 overflow, and (when I tested it) crashed. This was caused by some obsolete
3041 code that should have been removed. The fix is to remove it!
3043 3. When running in the test harness, delay a bit after writing a bounce
3044 message to get a bit more predictability in the log output.
3046 4. Added a call to search_tidyup() just before forking a reception process. In
3047 theory, someone could use a lookup in the expansion of smtp_accept_max_
3048 per_host which, without the tidyup, could leave open a database connection.
3050 5. Added the variables $recipient_data and $sender_data which get set from a
3051 lookup success in an ACL "recipients" or "senders" condition, or a router
3052 "senders" option, similar to $domain_data and $local_part_data.
3054 6. Moved the writing of debug_print from before to after the "senders" test
3057 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3058 problems for message scanning, either using a data ACL, or using
3059 local_scan() because the Received: header was not generated till after they
3060 were called (in order to set the time as the time of reception completion).
3061 I have revised the way this works. The header is now generated after the
3062 body is received, but before the ACL or local_scan() are called. After they
3063 are run, the timestamp in the header is updated.
3069 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3070 before starting a queue runner without re-exec. This happened only when
3071 deliver_drop_privilege was set or when the Exim user was set to root. The
3072 effect of the bug was that timeouts during subsequent deliveries caused
3073 crashes instead of being properly handled. The handler is now left at its
3074 default (and expected) setting.
3076 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3077 message, again when deliver_drop_privilege is set or Exim is run as root.
3078 The bug described in (1) was not present in this case, but the tidying up
3079 of the other signals was missing. I have made the two cases consistent.
3081 3. The ignore_target_hosts setting on a manualroute router was being ignored
3082 for hosts that were looked up using the /MX notation.
3084 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3087 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3088 operated on the sender address. After changing the $sender_address to <>
3089 for the sender address verify, Exim was re-instated it as the original
3090 (before rewriting) address, but remembering that it had rewritten it, so it
3091 wasn't rewriting it again. This bug also had the effect of breaking the
3092 sender address verification caching when the sender address was rewritten.
3094 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3095 This has been changed so that if the ip literal address matches
3096 ignore_target_hosts, the router declines.
3098 7. Added expansion conditions match_domain, match_address, and match_local_
3099 part (NOT match_host).
3101 8. The placeholder for the Received: header didn't have a length field set.
3103 9. Added code to Exim itself and to exim_lock to test for a specific race
3104 condition that could lead to file corruption when using MBX delivery. The
3105 issue is with the lockfile that is created in /tmp. If this file is removed
3106 after a process has opened it but before that process has acquired a lock,
3107 there is the potential for a second process to recreate the file and also
3108 acquire a lock. This could lead to two Exim processes writing to the file
3109 at the same time. The added code performs the same test as UW imapd; it
3110 checks after acquiring the lock that its file descriptor still refers to
3111 the same named file.
3113 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3114 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3117 11. Added the smtp_active_hostname option. If used, this will typically be made
3118 to depend on the incoming interface address. Because $interface_address is
3119 not set up until the daemon has forked a reception process, error responses
3120 that can happen earlier (such as "too many connections") no longer contain
3123 12. If an expansion in a condition on a "warn" statement fails because a lookup
3124 defers, the "warn" statement is abandoned, and the next ACL statement is
3125 processed. Previously this caused the whole ACL to be aborted.
3127 13. Added the iplsearch lookup type.
3129 14. Added ident_timeout as a log selector.
3131 15. Added tls_certificate_verified as a log selector.
3133 16. Added a global option tls_require_ciphers (compare the smtp transport
3134 option of the same name). This controls incoming TLS connections.
3136 17. I finally figured out how to make tls_require_ciphers do a similar thing
3137 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3138 before starting the TLS session.
3140 18. Tabs are now shown as \t in -bP output.
3142 19. If the log selector return_path_on_delivery was set, Exim crashed when
3143 bouncing a message because it had too many Received: header lines.
3145 20. If two routers both had headers_remove settings, and the first one included
3146 a superfluous trailing colon, the final name in the first list and the
3147 first name in the second list were incorrectly joined into one item (with a
3148 colon in the middle).
3154 1. Added -C and -D options to the exinext utility, mainly to make it easier
3155 to include in the automated testing, but these could be helpful when
3156 multiple configurations are in use.
3158 2. The exinext utility was not formatting the output nicely when there was
3159 an alternate port involved in the retry record key, nor when there was a
3160 message id as well (for retries that were specific to a specific message
3161 and a specific host). It was also confused by IPv6 addresses, because of
3162 the additional colons they contain. I have fixed the IPv4 problem, and
3163 patched it up to do a reasonable job for IPv6.
3165 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3166 delivery, the log line now contains "pipelined" if PIPELINING was used.
3168 4. An SMTP transport process used to panic and die if the bind() call to set
3169 an explicit outgoing interface failed. This has been changed; it is now
3170 treated in the same way as a connect() failure.
3172 5. A reference to $sender_host_name in the part of a conditional expansion
3173 that was being skipped was still causing a DNS lookup. This no longer
3176 6. The def: expansion condition was not recognizing references to header lines
3177 that used bh_ and bheader_.
3179 7. Added the _cache feature to named lists.
3181 8. The code for checking quota_filecount in the appendfile transport was
3182 allowing one more file than it should have been.
3184 9. For compatibility with Sendmail, the command line option
3192 and sets the incoming protocol and host name (for trusted callers). The
3193 host name and its colon can be omitted when only the protocol is to be set.
3194 Note the Exim already has two private options, -pd and -ps, that refer to
3195 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3196 "s", but I don't think that's a major issue.
3198 10. A number of refactoring changes to the code, none of which should affect
3201 (a) The number of logging options was getting close to filling up the
3202 32-bit word that was used as a bit map. I have split them into two classes:
3203 those that are passed in the argument to log_write(), and those that are
3204 only ever tested independently outside of that function. These are now in
3205 separate 32-bit words, so there is plenty of room for expansion again.
3206 There is no change in the user interface or the logging behaviour.
3208 (b) When building, for example, log lines, the code previously used a
3209 macro that called string_cat() twice, in order to add two strings. This is
3210 not really sufficiently general. Furthermore, there was one instance where
3211 it was actually wrong because one of the argument was used twice, and in
3212 one call a function was used. (As it happened, calling the function twice
3213 did not affect the overall behaviour.) The macro has been replaced by a
3214 function that can join an arbitrary number of extra strings onto a growing
3217 (c) The code for expansion conditions now uses a table and a binary chop
3218 instead of a serial search (which was left over from when there were very
3219 few conditions). Also, it now recognizes conditions like "pam" even when
3220 the relevant support is not compiled in: a suitably worded error message is
3221 given if an attempt is made to use such a condition.
3223 11. Added ${time_interval:xxxxx}.
3225 12. A bug was causing one of the ddress fields not to be passed back correctly
3226 from remote delivery subprocesses. The field in question was not being
3227 subsequently used, so this caused to problems in practice.
3229 13. Added new log selectors queue_time and deliver_time.
3231 14. Might have fixed a bug in maildirsizefile handling that threw up
3232 "unexpected character" debug warnings, and recalculated the data
3233 unnecessarily. In any case, I expanded the warning message to give more
3236 15. Added the message "Restricted characters in address" to the statements in
3237 the default ACL that block characters like @ and % in local parts.
3239 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3240 Three changes have been made:
3242 (a) There was a serious bug; a negative response to MAIL caused the whole
3243 recipient domain to be cached as invalid, thereby blocking all messages
3244 to all local parts at the same domain, from all senders. This bug has
3245 been fixed. The domain is no longer cached after a negative response to
3246 MAIL if the sender used is not empty.
3248 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3251 (c) A new callout option, "use_sender" has been added for people who want
3252 the modified behaviour.
3258 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3261 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3262 indeed breaks things for older releases.
3264 3. Added additional logging to the case where there is a problem reading data
3265 from a filter that is running in a subprocess using a pipe, in order to
3266 try to track down a specific problem.
3268 4. Testing facility fudge: when running in the test harness and attempting
3269 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3270 getting "No route to host". Convert this to a timeout.
3272 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3275 6. Some OS don't have socklen_t but use size_t instead. This affects the
3276 fifth argument of getsockopt() amongst other things. This is now
3277 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3278 can be set for individual OS. I have set it for SunOS5, OSF1, and
3279 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3280 some earlier ones do not.
3282 7. Change 4.30/15 was not doing the test caselessly.
3284 8. The standard form for an IPv6 address literal was being rejected by address
3285 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3286 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3287 this, as well as the form without the "IPv6" on the front (but only when
3288 address literals are enabled, of course).
3290 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3292 10. Exim crashed if a message with an empty sender address specified by -f
3293 encountered a router with an errors_to setting. This could be provoked only
3294 by a command such as
3298 where an empty string was supplied; "<>" did not hit this bug.
3300 11. Installed PCRE release 4.5.
3302 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3303 remained set. It is now erased.
3305 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3306 times from message ids (which are base 36 rather than the normal 62).
3308 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3309 were being counted as actual protocol errors, and logged if the log
3310 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3311 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3312 and DATA following a set of rejected RCPTs do not count as protocol errors.
3313 In other words, Exim assumes they were pipelined, though this may not
3314 actually be the case. Of course, in all cases the client gets an
3315 appropriate error code.
3317 15. If a lookup fails in an ACL condition, a message about the failure may
3318 be available; it is used if testing the ACL cannot continue, because most
3319 such messages specify what the cause of the deferral is. However, some
3320 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3321 that caused an old message to be retained and used if a later statement
3322 caused a defer, replacing the real cause of the deferral.
3324 16. If an IP address had so many PTR records that the DNS lookup buffer
3325 was not large enough to hold them, Exim could crash while trying to process
3326 the truncated data. It now detects and logs this case.
3328 17. Further to 4.21/58, another change has been made: if (and only if) the
3329 first line of a message (the first header line) ends with CRLF, a bare LF
3330 in a subsequent header line has a space inserted after it, so as not to
3331 terminate the header.
3333 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3334 unnecessarily, used atoi() instead of strtol(), and didn't check the
3335 termination when getting file sizes from file names by regex.
3337 19. Completely re-implemented the support for maildirsize files, in the light
3338 of a number of problems with the previous contributed implementation
3339 (4.30/29). In particular:
3341 . If the quota is zero, the maildirsize file is maintained, but no quota is
3344 . If the maildir directory does not exist, it is created before any attempt
3345 to write a maildirsize file.
3347 . The quota value in the file is just a cache; if the quota is changed in
3348 the transport, the new value overrides.
3350 . A regular expression is available for excluding directories from the
3353 20. The autoreply transport checks the characters in options that define the
3354 message's headers; it allows continued headers, but it was checking with
3355 isspace() after an embedded newline instead of explicitly looking for a
3358 21. If all the "regular" hosts to which an address was routed had passed their
3359 expiry times, and had not reached their retry times, the address was
3360 bounced, even if fallback hosts were defined. Now Exim should go on to try
3363 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3364 equivalent code in the SMTP transport. Some hosts send humungous responses
3365 to HELO/EHLO, more than 1024 it seems.
3367 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3368 gives compiler warnings in some environments. I've now done it "properly",
3371 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3372 (because of problems with the built-in one) was declared to return uschar *
3373 instead of char *, causing compiler failure.
3375 25. Fixed a file descriptor leak when processing alias/forward files.
3377 26. Fixed a minor format string issue in dbfn.c.
3379 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3381 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3382 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3385 29. When a sender address is verified, it is cached, to save repeating the test
3386 when there is more than one recipient in a message. However, when the
3387 verification involves a callout, it is possible for different callout
3388 options to be set for different recipients. It is too complicated to keep
3389 track of this in the cache, so now Exim always runs a verification when a
3390 callout is required, relying on the callout cache for the optimization.
3391 The overhead is duplication of the address routing, but this should not be
3394 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3395 to be verified with callout=postmaster, and the main callout worked but the
3396 postmaster check failed, the verification correctly failed. However, if a
3397 subsequent RCPT command asked for sender verification *without* the
3398 postmaster check, incorrect caching caused this verification also to fail,
3401 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3402 it was not caching the DNS options (qualify_single, search_parents) that
3403 were used when the lookup failed. A subsequent lookup with different
3404 options therefore always gave the same answer, though there were cases
3405 where it should not have. (Example: a "domains = !$mx_any" option on a
3406 dnslookup router: the "domains" option is always processed without any
3407 widening, but the router might have qualify_single set.) Now Exim uses the
3408 cached value only when the same options are set.
3410 32. Added John Jetmore's "exipick" utility to the distribution.
3412 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3413 than a timeout (e.g. a certificate is required, and is not provided), an
3414 Exim server now closes the connection immediately. Previously it waited for
3415 the client to close - but if the client is SSL, it seems that they each
3416 wait for each other, leading to a delay before one of them times out.
3418 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3419 maintained 0.8.x compatibility because I don't think many are using it, and
3420 it is clearly obsolete.
3422 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3425 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3426 client certificate was expired. A simple patch fixes this, though I don't
3427 understand the full logic of why the verify callback is called multiple
3430 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3431 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3432 which causes problems with some clients (such as the Certicom SSL Plus
3433 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3434 disables the coutermeasure allowing Eudora to connect."
3436 38. Exim was not checking that a write() to a log file succeeded. This could
3437 lead to Bad Things if a log got too big, in particular if it hit a file
3438 size limit. Exim now panics and dies if it cannot write to a log file, just
3439 as it does if it cannot open a log file.
3441 39. Modified OS/Makefile-Linux so that it now contains
3443 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3445 The two -D definitions ensure that Exim is compiled with large file
3446 support, which makes it possible to handle log files that are bigger than
3449 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3450 instance) a domain was checked against a named list that involved a lookup,
3451 causing $domain_data to be set, then another domain was checked against the
3452 same list, then the first domain was re-checked, the value of $domain_data
3453 after the final check could be wrong. In particular, if the second check
3454 failed, it could be set empty. This bug probably also applied to
3457 41. The strip_trailing_dot option was not being applied to the address given
3458 with the -f command-line option.
3460 42. The code for reading a message's header from the spool was incrementing
3461 $received_count, but never initializing it. This meant that the value was
3462 incorrect (doubled) while delivering a message in the same process in which
3463 it was received. In the most common configuration of Exim, this never
3464 happens - a fresh exec is done - but it can happen when
3465 deliver_drop_privilege is set.
3467 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3468 it now includes up to 150 characters of the unexpected data in the log
3471 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3472 and building data strings. The size of both of these buffers was 10 000
3473 bytes - far larger than anybody would *ever* want, thought I. Needless to
3474 say, somebody hit the limit. I have increased the maximum line length to
3475 20 000 and the maximum data length of concatenated lines to 100 000. I have
3476 also fixed two bugs, because there was no checking on these buffers. Tsk,
3477 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3478 buffer is too small.
3480 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3481 lsearch lookups. Now it does.
3483 46. When parsing a route_list item in a manualroute router, a fixed-length
3484 buffer was used for the list of hosts. I made this 1024 bytes long,
3485 thinking that nobody would ever have a list of hosts that long. Wrong.
3486 Somebody had a whole pile of complicated expansion conditions, and the
3487 string was silently truncated, leading to an expansion error. It turns out
3488 that it is easier to change to an unlimited length (owing to other changes
3489 that have happened since this code was originally written) than to build
3490 structure for giving a limitation error. The length of the item that
3491 expands into the list of hosts is now unlimited.
3493 47. The lsearch lookup could not handle data where the length of text line was
3494 more than 4095 characters. Such lines were truncated, leading to shortened
3495 data being returned. It should now handle lines of any length.
3497 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3498 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3501 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3502 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3503 no longer happens in comment lines. A list of replacements is now placed
3504 at the head of all of the source files, except those whose only change is
3505 to replace PERL_COMMAND in the very first #! line.
3507 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3508 messages on the queue, with a bottom-up merge sort, using code contributed
3509 by Michael Haardt. This should make operations like -bp somewhat faster on
3510 large queues. It won't affect queue runners, except when queue_run_in_order
3513 51. Installed eximstats 1.31 in the distribution.
3515 52. Added support for SRV lookups to the dnslookup router.
3517 53. If an ACL referred to $message_body or $message_body_end, the value was not
3518 reset for any messages that followed in the same SMTP session.
3520 54. The store-handling optimization for building very long strings was not
3521 differentiating between the different store pools. I don't think this
3522 actually made any difference in practice, but I've tidied it.
3524 55. While running the routers to verify a sender address, $sender_address
3525 was still set to the sender address. This is wrong, because when routing to
3526 send a bounce to the sender, it would be empty. Therefore, I have changed
3527 it so that, while verifying a sender address, $sender_address is set to <>.
3528 (There is no change to what happens when verifying a recipient address.)
3530 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3531 target A or AAAA records (if not already returned) without resetting the
3532 qualify_single or search_parents options of the DNS resolver. These are
3533 inappropriate in this case because the targets of MX and SRV records must
3534 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3535 target that, when qualified, matched something in the local domain. These
3536 two options are now turned off when doing these lookups.
3538 57. It seems that at least some releases of Reiserfs (which does not have the
3539 concept of a fixed number of inodes) returns zero and not -1 for the
3540 number of available inodes. This interacted badly with check_spool_inodes,
3541 which assumed that -1 was the "no such thing" setting. What I have done is
3542 to check that the total number of inodes is greater than zero before doing
3543 the test of how many are available.
3545 58. When a "warn" ACL statement has a log_message modifier, the message is
3546 remembered, and not repeated. This is to avoid a lot of repetition when a
3547 message has many recipients that cause the same warning to be written.
3548 Howewer, Exim was preserving the list of already written lines for an
3549 entire SMTP session, which doesn't seem right. The memory is now reset if a
3550 new message is started.
3552 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3553 debugging output unless log_rewrite was also set.
3555 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3556 of (int)(handle) when we know that handle contains (void *)(-1).
3558 61. The Exim daemon panic-logs an error return when it closes the incoming
3559 connection. However "connection reset by peer" seems to be common, and
3560 isn't really an error worthy of noting specially, so that particular error
3563 62. When Exim is trying to find all the local interfaces, it used to panic and
3564 die if the ioctl to get the interface flags failed. However, it seems that
3565 on at least one OS (Solaris 9) it is possible to have an interface that is
3566 included in the list of interfaces, but for which you get a failure error
3567 for this call. This happens when the interface is not "plumbed" into a
3568 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3569 failure of the "get flags" call assumes that the interface is down.
3571 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3572 makes life easier for people who are doing arithmetic on fields extracted
3573 from dates, where you often get leading zeros that should not be
3574 interpreted as octal.
3576 64. Added qualify_domain to the redirect router, to override the global
3579 65. If a pathologically long header line contained very many addresses (the
3580 report of this problem mentioned 10 000) and each of them was rewritten,
3581 Exim could use up a very large amount of memory. (It kept on making new
3582 copies of the header line as it rewrote, and never released the old ones.)
3583 At the expense of a bit more processing, the header rewriting function has
3584 been changed so that it no longer eats memory in this way.
3586 66. The generation of the Received: header has been moved from the time that a
3587 message starts to be received, to the time that it finishes. The timestamp
3588 in the Received: header should now be very close to that of the <= log
3589 line. There are two side-effects of this change:
3591 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3592 logged header lines no longer include the local Received: line, because
3593 it has not yet been created. The same applies to a copy of the message
3594 that is returned to a non-SMTP sender when a message is rejected.
3596 (b) When a filter file is tested using -bf, no additional Received: header
3597 is added to the test message. After some thought, I decided that this
3600 This change does not affect the value of $received_for. It is still set
3601 after address rewriting, but before local_scan() is called.
3603 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3605 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3606 gave an unhelpful panic error message, and a defer error. I have managed to
3607 change this behaviour so that it now rejects any supplied certificate,
3608 which seems right, as the list of acceptable certificates is empty.
3610 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3611 gave an unhelpful defer error. I have not managed to make this reject any
3612 supplied certificates, but the error message it gives is "no certificate
3613 supplied", which is not helpful.
3615 70. exigrep's output now also includes lines that are not associated with any
3616 message, but which match the given pattern. Implemented by a patch from
3617 Martin Sluka, which also tidied up the Perl a bit.
3619 71. Recipient callout verification, like sender verification, was using <> in
3620 the MAIL FROM command. This isn't really the right thing, since the actual
3621 sender may affect whether the remote host accepts the recipient or not. I
3622 have changed it to use the actual sender in the callout; this means that
3623 the cache record is now keyed on a recipient/sender pair, not just the
3624 recipient address. There doesn't seem to be a real danger of callout loops,
3625 since a callout by the remote host to check the sender would use <>.
3626 [SEE ABOVE: changed after hitting problems.]
3628 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3629 temporary errors. However, in the case of such a code being given after
3630 the end of a data transmission (i.e. after ".") Exim was failing to write
3631 a retry record for the message. (Yes, there was some broken host that was
3632 actually sending 8xx at this point.)
3634 73. An unknown lookup type in a host list could cause Exim to panic-die when
3635 the list was checked. (An example that provoked this was putting <; in the
3636 middle of a list instead of at the start.) If this happened during a DATA
3637 ACL check, a -D file could be left lying around. This kind of configuration
3638 error no longer causes Exim to die; instead it causes a defer errror. The
3639 incident is still logged to the main and panic logs.
3641 74. Buglet left over from Exim 3 conversion. The message "too many messages
3642 in one connection" was written to the rejectlog but not the mainlog, except
3643 when address rewriting (yes!) was being logged.
3645 75. Added write_rejectlog option.
3647 76. When a system filter was run not as root (that is, when system_filter_user
3648 was set), the values of the $n variables were not being returned to the
3649 main process; thus, they were not subsequently available in the $sn
3652 77. Added +return_path_on_delivery log selector.
3654 78. A connection timeout was being treated differently from recipients deferred
3655 when testing hosts_max_try with a message that was older than the host's
3656 retry timeout. (The host should not be counted, thus allowing all hosts to
3657 be tried at least once before bouncing.) This may have been the cause of an
3658 occasionally reported bug whereby a message would remain on the queue
3659 longer than the retry timeout, but would be bounced if a delivery was
3660 forced. I say "may" because I never totally pinned down the problem;
3661 setting up timeout/retry tests is difficult. See also the next item.
3663 79. The ultimate address timeout was not being applied to errors that involved
3664 a combination of host plus message (for example, a timeout on a MAIL
3665 command). When an address resolved to a number of possible hosts, and they
3666 were not all tried for each delivery (e.g. because of hosts_max_try), a
3667 message could remain on the queue longer than the retry timeout.
3669 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3672 81. Fixed an obscure SMTP outgoing bug which required at least the following
3673 conditions: (a) there was another message waiting for the same server;
3674 (b) the server returned 5xx to all RCPT commands in the first message so
3675 that the message was not completed; (c) the server dropped the connection
3676 or gave a negative response to the RSET that Exim sends to abort the
3677 transaction. The observed case was a dropped connection after DATA that had
3678 been sent in pipelining mode. That is, the server had advertised PIPELINING
3679 but was not implementing it correctly. The effect of the bug was incorrect
3680 behaviour, such as trying another host, and this could lead to a crash.
3686 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3687 and daemon.c were passed as pointers to ints; they should have been
3688 pointers to socklen_t variables (which are typically unsigned ints).
3690 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3693 3. Fixed a really odd bug that affected only the testing scheme; patching a
3694 certain fixed string in the binary changed the value of another string that
3695 happened to be identical to the end of the original first string.
3697 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3698 name", it returns that address as the IP address. On some operating
3699 systems (e.g. Solaris), it also passes back the IP address string as the
3700 "host name". However, on others (e.g. Linux), it passes back an empty
3701 string. Exim wasn't checking for this, and was changing the host name to an
3702 empty string, assuming it had been canonicized.
3704 5. Although rare, it is permitted to have more than one PTR record for a given
3705 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3706 all the names associated with an address, because they do in Solaris.
3707 However, it seems that they do not in Linux for data that comes from the
3708 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3709 I found this out when I moved to a new Linux workstation and tried to run
3710 the Exim test suite.
3712 To get round this problem I have changed the code so that it now does its
3713 own call to the DNS to look up PTR records when searching for a host name.
3714 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3715 addresses that are only in /etc/hosts are still found.
3717 This behaviour is, however, controlled by an option called host_lookup_
3718 order, which defaults to "bydns:byaddr". If people want to use the other
3719 order, or indeed, just use one or the other means of lookup, they can
3720 specify it in this variable.
3722 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3723 some operating systems, this comes back from gethostbyaddr() as an empty
3724 string, and this is what Exim used to test for. However, it seems that in
3725 other systems, "." is yielded. Exim now tests for this case too.
3727 7. The values of check_spool_space and check_log_space are now held internally
3728 as a number of kilobytes instead of an absolute number of bytes. If a
3729 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3730 kilobyte. This means that much larger values can be stored.
3732 8. Exim monitor: an attempt to get the action menu when not actually pointing
3733 at a message produces an empty menu entitled "No message selected". This
3734 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3735 no entries in it ("Shell widget menu has zero width and/or height"). So I
3736 have added a single, blank menu entry in this case.
3738 9. Added ${quote_local_part.
3740 10. MIME decoding is now applied to the contents of Subject: header lines when
3743 11. Now that a reference to $sender_host_address automatically causes a reverse
3744 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3745 host lookup before query-style lookups in lists that might use this
3746 variable. This has therefore been abolished, and the "net-" prefix is no
3747 longer necessary for query-style lookups.
3749 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3750 have been a typo for LFLAGS, so it has been changed.
3752 13. The install script calls Exim with "-C /dev/null" in order to find the
3753 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3754 to be output. Howeve, since Exim outputs its version number before the
3755 error, it didn't break the script. It just looked ugly. I fixed this by
3756 always allowing "-C /dev/null" if the caller is root.
3758 14. Ignore overlarge ACL variable number when reading spool file - insurance
3759 against a later release with more variables having written the file.
3761 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3762 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3763 this, as well as the form without the "IPv6" on the front.
3765 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3766 OS/Makefile-Darwin file.
3768 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3769 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3771 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3772 variable. However, some OS use ints for inodes. I've added cast to long int
3773 to get rid of the compiler warning.
3775 19. I had forgotten to lock out "/../" in configuration file names when
3776 ALT_CONFIG_PREFIX was set.
3778 20. Routers used for verification do not need to specify transports. However,
3779 if such a router generated a host list, and callout was configured, Exim
3780 crashed, because it could not find a port number from the (non-existent)
3781 transport. It now assumes port 25 in this circumstance.
3783 21. Added the -t option to exigrep.
3785 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3786 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3787 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3788 them is set, LOOKUP_LSEARCH is forced.
3790 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3792 24. Added sender and host information to the "rejected by local_scan()" log
3793 line; previously there was no indication of these.
3795 25. Added .include_if_exists.
3797 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3798 Linux. It turns out that not all file systems support this. Apparently some
3799 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3800 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3801 not supported on the file descriptor, is now ignored when Exim is trying to
3802 sync a directory. This applies only to Linux.
3804 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3806 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3807 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3810 29. Added support for maildirsize files from supplied patch (modified a bit).
3812 30. The use of :fail: followed by an empty string could lead Exim to respond to
3813 sender verification failures with (e.g.):
3815 550 Verification failed for <xxx>
3816 550 Sender verify failed
3818 where the first response line was missing the '-' that indicates it is not
3819 the final line of the response.
3821 31. The loop for finding the name of the user that called Exim had a hardwired
3822 limit of 10; it now uses the value of finduser_retries, which is used for
3823 all other user lookups.
3825 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3828 33. Exim was neglecting to zero errno before one call of strtol() when
3829 expanding a string and expecting an integer value. On some systems this
3830 resulted in spurious "integer overflow" errors. Also, it was casting the
3831 result into an int without checking.
3833 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3834 did not work. The code looks as if it has *never* worked, though it appears
3835 to have been documented since at least releast 1.62. I have made it work.
3837 35. The "timeout_DNS" error in retry rules, also documented since at least
3838 1.62, also never worked. As it isn't clear exactly what this means, and
3839 clearly it isn't a major issue, I have abolished the feature by treating it
3840 as "timeout", and writing a warning to the main and panic logs.
3842 36. The display of retry rules for -brt wasn't always showing the error code
3845 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3846 timeout_connect_A, timeout_connect_MX.
3848 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3849 to the empty sender.
3851 39. The daemon was not analysing the content of -oX till after it had closed
3852 stderr and disconnected from the controlling terminal. This meant that any
3853 syntax errors were only noted on the panic log, and the return code from
3854 the command was 0. By re-arranging the code a little, I've made the
3855 decoding happen first, so such errors now appear on stderr, and the return
3856 code is 1. However, the actual setting up of the sockets still happens in
3857 the disconnected process, so errors there are still only recorded on the
3860 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3861 connections (as happens on some IP stacks) was logged at start up time as
3862 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3863 it from "IPv6 and IPv4", which means that two separate sockets are being
3866 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3867 says whether AF_INET or AF_INET6 was passed as an argument.
3869 42. Exiwhat output was messed up when time zones were included in log
3872 43. Exiwhat now gives more information about the daemon's listening ports,
3873 and whether -tls-on-connect was used.
3875 44. The "port" option of the smtp transport is now expanded.
3877 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3878 silently ignored. Now an error message is written to the main and panic
3881 46. There's a new ACL modifier called "logwrite" which writes to a log file
3882 as soon as it is encountered.
3884 47. Added $local_user_uid and $local_user_gid at routing time.
3886 48. Exim crashed when trying to verify a sender address that was being
3889 49. Exim was recognizing only a space character after ".include". It now also
3890 recognizes a tab character.
3892 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3893 extracting the relevant information from the specification. The man page no
3894 longer contains scrambled data for the -d option, and I've added a section
3895 at the front about calling Exim under different names.
3897 51. Added "extra_headers" argument to the "mail" command in filter files.
3899 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3902 53. Installed eximstats 1.29.
3904 54. Added transport_filter_timeout as a generic transport option.
3906 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3907 Cc: header lines. This was required by RFC 822, but it not required by RFC
3910 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3911 incoming messages that did not have them. Now it does so only if the
3912 message originates locally, that is, if there is no associated remote host
3913 address. When Resent- header lines are present, this applies to the Resent-
3914 lines rather than the non-Resent- lines.
3916 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3917 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3919 58. Messages for configuration errors now include the name of the main
3920 configuration file - useful now that there may be more than one file in a
3921 list (.included file names were always shown).
3923 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3924 for those rare installations that do not start the daemon as root or run it
3925 setuid root. I've cut out the call to initgroups() if the daemon is not
3928 60. The Exim user and group can now be bound into the binary as text strings
3929 that are looked up at the start of Exim's processing.
3931 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3933 62. Added $mailstore_basename variable.
3935 63. Installed patch to sieve.c from Michael Haardt.
3937 64. When Exim failed to open the panic log after failing to open the main log,
3938 the original message it was trying to log was written to stderr and debug
3939 output, but if they were not available (the usual case in production), it
3940 was lost. Now it is written to syslog before the two lines that record the
3941 failures to open the logs.
3943 65. Users' Exim filters run in subprocesses under the user's uid. It is
3944 possible for a "deliver" command or an alias in a "personal" command to
3945 provoke an address rewrite. If logging of address rewriting is configured,
3946 this fails because the process is not running as root or exim. There may be
3947 a better way of dealing with this, but for the moment (because 4.30 needs
3948 to be released), I have disabled address rewrite logging when running a
3949 filter in a non-root, non-exim process.
3955 1. The buildconfig auxiliary program wasn't quoting the value set for
3956 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3957 not defined. This bug was masked in 4.22 by the effect that was fixed in
3960 2. Some messages that were rejected after a message id was allocated were
3961 shown as "incomplete" by exigrep. It no longer does this for messages that
3962 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3964 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3965 have allow_domain_literals set, the ID did not get logged in the <= line.
3966 Domain literals are now always recognized in Message-ID: header lines.
3968 4. The first argument for a ${extract expansion item is the key name or field
3969 number. Leading and trailing spaces in this item were not being ignored,
3970 causing some misleading effects.
3972 5. When deliver_drop_privilege was set, single queue runner processes started
3973 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3974 same command in the process it spins off) were not dropping privilege.
3976 6. When the daemon running as "exim" started a queue runner, it always
3977 re-executed Exim in the spun-off process. This is a waste of effort when
3978 deliver_drop_privilege is set. The new process now just calls the
3979 queue-runner function directly.
3985 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3988 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3989 ignored. Though the use of -oP was forcing the writing of a pid file, it
3990 was always written to the default place.
3992 3. If the message "no IP address found for host xxxx" is generated during
3993 incoming verification, it is now followed by identification of the incoming
3994 connection (so you can more easily find what provoked it).
3996 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3998 5. Added some features to "harden" Exim a bit more against certain attacks:
4000 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4001 be put in Local/Makefile. This is like the never_users runtime option,
4002 but it cannot be overridden. The default setting is "root".
4004 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4005 prefix string with which any file named in a -C command line option
4008 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4009 is retained for -C and -D only if the caller of Exim is root. Without
4010 it, the exim user may also use -C and -D and retain privilege.
4012 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4013 command line option is disabled.
4015 6. Macro names set by the -D option must start with an upper case letter, just
4016 like macro names defined in the configuration file.
4018 7. Added "dereference=" facility to LDAP.
4020 8. Two instances of the typo "uknown" in the source files are fixed.
4022 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4023 the Configure-Makefile script screwed up while processing it.
4025 10. Incorporated PCRE 4.4.
4027 11. The SMTP synchronization check was not operating right at the start of an
4028 SMTP session. For example, it could not catch a HELO sent before the client
4029 waited for the greeting. There is now a check for outstanding input at the
4030 point when the greeting is written. Because of the duplex, asynchronous
4031 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4032 way, but not yet received, when the check is performed.
4034 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4035 on TCP/IP sockets, because this apparently causes some broken clients to
4038 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4039 unchanged) from the Cygwin maintainer.
4041 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4042 was supported instead of testing for mbx. Effectively a typo.
4044 15. The spa authenticator server code was not checking that the input it
4045 received was valid base64.
4047 16. The debug output line for the "set" modifier in ACLs was not showing the
4048 name of the variable that was being set.
4050 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4052 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4053 lookup on its own. Something else was needed to trigger the lookup. For
4054 example, a match in host_lookup or the need for a host name in a host list.
4055 Now, if $sender_host_name is referenced and the host name has not yet been
4056 looked up, a lookup is performed. If the lookup fails, the variable remains
4057 empty, and $host_lookup_failed is set to "1".
4059 19. Added "eqi" as a case-independent comparison operator.
4061 20. The saslauthd authentication condition could segfault if neither service
4062 nor realm was specified.
4064 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4065 error message that was logged was misleading, and incoming SMTP
4066 connections were dropped. The message is now more accurate, and temporary
4067 errors are given to SMTP connections.
4069 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4070 (except RSET) with a 421 error, until QUIT is received. However, it was
4071 failing to send a response to QUIT.
4073 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4074 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4075 ACL is now run after the helo_try_verify_hosts code.
4077 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4078 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4079 case-independent, so other case variants are also recognized). Apparently
4080 some systems use these upper case variants.
4082 25. If more than two messages were waiting for the same host, and a transport
4083 filter was specified for the transport, Exim sent two messages over the
4084 same TCP/IP connection, and then failed with "socket operation on non-
4085 socket" when it tried to send the third.
4087 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4089 27. The extern definition of crypt16() in expand.c was not being excluded when
4090 the OS had its own crypt16() function.
4092 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4093 as a preferred synonym for return_size_limit, both as an option and as an
4096 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4098 30. Changed the default configuration ACL to relax the local part checking rule
4099 for addresses that are not in any local domains. For these addresses,
4100 slashes and pipe symbols are allowed within local parts, but the sequence
4101 /../ is explicitly forbidden.
4103 31. SPA server authentication was not clearing the challenge buffer before
4106 32. log_message in a "warn" ACL statement was writing to the reject log as
4107 well as to the main log, which contradicts the documentation and doesn't
4108 seem right (because no rejection is happening). So I have stopped it.
4110 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4111 However, I am unable to do any testing of this.
4113 34. Fixed an infelicity in the appendfile transport. When checking directories
4114 for a mailbox, to see if any needed to be created, it was accidentally
4115 using path names with one or more superfluous leading slashes; tracing
4116 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4118 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4119 contents are added to the log line that is written for every discarded
4120 recipient. (Previously a log_message setting was ignored.)
4122 36. The ${quote: operator now quotes the string if it is empty.
4124 37. The install script runs exim in order to find its version number. If for
4125 some reason other than non-existence or emptiness, which it checks, it
4126 could not run './exim', it was installing it with an empty version number,
4127 i.e. as "exim-". This error state is now caught, and the installation is
4130 38. An argument was missing from the function that creates an error message
4131 when Exim fails to connect to the socket for saslauthd authentication.
4132 This could cause Exim to crash, or give a corrupted message.
4134 39. Added isip, isip4, and isip6 to ${if conditions.
4136 40. The ACL variables $acl_xx are now saved with the message, and can be
4137 accessed later in routers, transports, and filters.
4139 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4140 strings in the file are not string-expanded.
4142 42. If a MAIL command specified a SIZE value that was too large to fit into an
4143 int variable, the check against message_size_limit failed. Such values are
4144 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4145 day this will have to be increased, but I don't think I want to be around
4146 when emails are that large.
4153 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4154 iconv() is not standard in FreeBSD.
4156 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4157 IPv6 enabled. The observed symptom was a segmentation fault on return from
4158 the function os_common_find_running_interfaces() in src/os.c.
4160 3. In the check_special_case() function in daemon.c I had used "errno" as an
4161 argument name, which causes warnings on some systems. This was basically a
4162 typo, since it was named "eno" in the comments!
4164 4. The code that waits for the clock to tick (at a resolution of some fraction
4165 of a second) so as to ensure message-id uniqueness was always waiting for
4166 at least one whole tick, when it could have waited for less. [This is
4167 almost certainly not relevant at current processor speeds, where it is
4168 unlikely to ever wait at all. But we try to future-proof.]
4170 5. The function that sleeps for a time interval that includes fractions of a
4171 second contained a race. It did not block SIGALRM between setting the
4172 timer, and suspending (a couple of lines later). If the interval was short
4173 and the sigsuspend() was delayed until after it had expired, the suspension
4174 never ended. On busy systems this could lead to processes getting stuck for
4177 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4178 process, before it forks any delivery processes. The open lookup caching
4179 mechanism meant that the open file or database connection was passed into
4180 the delivery process. The problem was that delivery processes always tidy
4181 up cached lookup data. This could cause a problem for the next delivery
4182 process started by the queue runner, because the external queue runner
4183 process does not know about the closure. So the next delivery process
4184 still has data in the lookup cache. In the case of a file lookup, there was
4185 no problem because closing a file descriptor in a subprocess doesn't affect
4186 the parent. However, if the lookup was caching a connection to a database,
4187 the connection was closed, and the second delivery process was likely to
4188 see errors such as "PGSQL: query failed: server closed the connection
4189 unexpectedly". The problem has been fixed by closing all cached lookups
4190 in a queue runner before running a delivery process.
4192 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4193 seem to have the "const" qualifier which it has on other OS. I've
4196 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4197 *of the same type* (client or server) with the same public name that an
4198 error should be diagnosed.
4200 9. When Exim looked up a host name for an IP address, but failed to find the
4201 original IP address when looking up the host name (a safety check), it
4202 output the message "<ip address> does not match any IP for NULL", which was
4203 confusing, to say the least. The bug was that the host name should have
4204 appeared instead of "NULL".
4206 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4207 user that is built into the binary, and the -C or -D options is used, root
4208 privilege is dropped before the configuration file is read. In addition,
4209 logging is switched to stderr instead of the normal log files. If the
4210 configuration then re-defines the Exim user, the unprivileged environment
4211 is probably not what is expected, so Exim logs a panic warning message (but
4214 However, if deliver_drop_privilege is set, the unprivileged state may well
4215 be exactly what is intended, so the warning has been cut out in that case,
4216 and Exim is allowed to try to write to its normal log files.
4222 1. smtp_return_error_details was not giving details for temporary sender
4223 or receiver verification errors.
4225 2. Diagnose a configuration error if two authenticators have the same public
4228 3. Exim used not to create the message log file for a message until the first
4229 delivery attempt. This could be confusing when incoming messages were held
4230 for policy or load reasons. The message log file is now created at the time
4231 the message is received, and an initial "Received" line is written to it.
4233 4. The automatically generated man page for command line options had a minor
4234 bug that caused no ill effects; however, a more serious problem was that
4235 the procedure for building the man page automatically didn't always
4236 operate. Consequently, release 4.20 contains an out-of-date version. This
4237 shouldn't happen again.
4239 5. When building Exim with embedded Perl support, the script that builds the
4240 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4241 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4243 6. The freeze_tell option was not being used for messages that were frozen on
4244 arrival, either by an ACL or by local_scan().
4246 7. Added the smtp_incomplete_transaction log selector.
4248 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4249 was accepting AUTH without a new EHLO.
4251 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4252 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4255 10. Logging of TCP/IP connections (when configured) now happens in the main
4256 daemon process instead of the child process, so that the TCP/IP connection
4257 count is more accurate (but it can never be perfect).
4259 11. The use of "drop" in a nested ACL was not being handled correctly in the
4260 outer ACL. Now, if condition failure induced by the nested "drop" causes
4261 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4262 or "require"), the connection is dropped.
4264 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4265 that yield "discard" can now be used with an "accept" or a "discard" verb,
4266 but an error is generated for any others (because I can't see a useful way
4267 to define what should happen).
4269 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4270 are now processed in the same way as lines in the Exim configuration file.
4271 In particular, continuation lines are supported.
4273 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4275 15. Added -ti meaning -t -i.
4277 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4278 domains, and warn by logging if others are found.
4280 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4281 returned from GIFCONF when Exim is trying to find the list of interfaces on
4282 a host. The code in os.c has been modified to copy each ifreq to an aligned
4283 structure in all cases.
4285 Also, in some cases, the returned ifreq's were being copied to a 'struct
4286 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4287 means the last couple of bytes of an IPv6 address could be chopped if the
4288 ifreq contained only a normal sockaddr (14 bytes storage).
4290 18. Named domain lists were not supported in the hosts_treat_as_local option.
4291 An entry such as +xxxx was not recognized, and was treated as a literal
4294 19. Ensure that header lines added by a DATA ACL are included in the reject log
4295 if the ACL subsequently rejects the message.
4297 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4298 MD5 (which is deprecated).
4300 21. When testing a filter file using -bf, Exim was writing a message when it
4301 took the sender from a "From " line in the message, but it was not doing so
4302 when it took $return_path from a Return-Path: header line. It now does.
4304 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4305 with a valid header line field name (a series of printing characters
4306 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4308 23. Changed "disc" in the source to "disk" to conform to the documentation and
4309 the book and for uniformity.
4311 24. Ignore Sendmail's -Ooption=value command line item.
4313 25. When execve() failed while trying to run a command in a pipe transport,
4314 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4315 could be confused with a return value of 69 from the command itself. This
4316 has been changed to 127, the value the shell returns if it is asked to run
4317 a non-existent command. The wording for the related log line suggests a
4318 non-existent command as the problem.
4320 26. If received_header_text expands to an empty string, do not add a Received:
4321 header line to the message. (Well, it adds a token one on the spool, but
4322 marks it "old" so that it doesn't get used or transmitted.)
4324 27. Installed eximstats 1.28 (addition of -nt option).
4326 28. There was no check for failure on the call to getsockname() in the daemon
4327 code. This can fail if there is a shortage of resources on the system, with
4328 ENOMEM, for example. A temporary error is now given on failure.
4330 29. Contrary to the C standard, it seems that in some environments, the
4331 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4332 program. Exim now does this explicitly; it affects the formatting of
4333 timestamps using strftime().
4335 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4336 complaints. I've now initialized all the variables, to avoid this.
4338 32. Header lines added by a system filter were not being "seen" during
4339 transport-time rewrites.
4341 33. The info_callback() function passed to OpenSSL is set up with type void
4342 (*)(SSL *, int, int), as described somewhere. However, when calling the
4343 function (actually a macro) that sets it up, the type void(*)() is
4344 expected. I've put in a cast to prevent warnings from picky compilers.
4346 34. If a DNS black list lookup found a CNAME record, but there were no A
4347 records associated with the domain it pointed at, Exim crashed.
4349 35. If a DNS black list lookup returned more than one A record, Exim ignored
4350 all but the first. It now scans all returned addresses if a particular IP
4351 value is being sought. In this situation, the contents of the
4352 $dnslist_value variable are a list of all the addresses, separated by a
4355 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4356 to accept a host name and all its aliases if the forward lookup for any of
4357 them yielded the IP address of the incoming connection. Now it accepts only
4358 those names whose forward lookup yields the correct IP address. Any other
4359 names are discarded. This closes a loophole whereby a rogue DNS
4360 administrator could create reverse DNS records to break through a
4361 wildcarded host restriction in an ACL.
4363 37. If a user filter or a system filter that ran in a subprocess used any of
4364 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4365 the wrong values were passed to the pipe command ($thisaddress had the
4366 value of $0, $0 had the value of $1, etc). This bug was introduced by
4367 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4369 38. Improved the line breaking for long SMTP error messages from ACLs.
4370 Previously, if there was no break point between 40 and 75 characters, Exim
4371 left the rest of the message alone. Two changes have been made: (a) I've
4372 reduced the minimum length to 35 characters; (b) if it can't find a break
4373 point between 35 and 75 characters, it looks ahead and uses the first one
4374 that it finds. This may give the occasional overlong line, but at least the
4375 remaining text gets split now.
4377 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4378 file descriptors might be low, and that setting 1000 would always raise it.
4379 It turns out that in some environments, the limit is already over 1000 and
4380 that lowering it causes trouble. So now Exim takes care not to decrease it.
4382 40. When delivering a message, the value of $return_path is set to $sender_
4383 address at the start of routing (routers may change the value). By an
4384 oversight, this default was not being set up when an address was tested by
4385 -bt or -bv, which affected the outcome if any router or filter referred to
4388 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4389 log only when "message" or "log_message" are set. However, if one of the
4390 conditions was an address verification, or a call to a nested ACL, the
4391 messages generated by the underlying test were being passed through. This
4392 no longer happens. The underlying message is available in $acl_verify_
4393 message for both "message" and "log_message" expansions, so it can be
4394 passed through if needed.
4396 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4397 new expansion $bh_ to give the encoded byte string without charset
4398 translation. Translation happens only if iconv() is available; HAVE_ICONV
4399 indicates this at build time. HEADERS_CHARSET gives the charset to
4400 translate to; headers_charset can change it in the configuration, and
4401 "headers charset" can change it in an individual filter file.
4403 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4404 that creates RFC 2047 encoded "words" labels them as that charset instead
4405 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4406 expansion operator; (ii) when Exim creates a From: line for a local
4407 message; (iii) when a header line is rewritten to include a "phrase" part.
4409 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4410 buggy, causing it to skip the first lines of messages whose message ID
4411 ended in 'D'. This would not have bitten before Exim release 4.14, because
4412 message IDs were unlikely to end in 'D' before then. The effect was to have
4413 incorrect size information for certain domains.
4415 45. #include "config.h" was missing at the start of the crypt16.c module. This
4416 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4419 46. If there was a timeout during a "random" callout check, Exim treated it as
4420 a failure of the random address, and carried on sending RSET and the real
4421 address. If the delay was just some slowness somewhere, the response to the
4422 original RCPT would be taken as a response to RSET and so on, causing
4423 mayhem of various kinds.
4425 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4426 when I implemented it. It didn't allow for the fact that some option values
4427 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4428 the right test for positive values.
4430 48. Domain names in DNS records are case-independent. Exim always looks them up
4431 in lower case. Some resolvers return domain names in exactly the case they
4432 appear in the zone file, that is, they may contain uppercase letters. Not
4433 all resolvers do this - some return always lower case. Exim was treating a
4434 change of case by a resolver as a change of domain, similar to a widening
4435 of a domain abbreviation. This triggered its re-routing code and so it was
4436 trying to route what was effectively the same domain again. This normally
4437 caused routing to fail (because the router wouldn't handle the domain
4438 twice). Now Exim checks for this case specially, and just changes the
4439 casing of the domain that it ultimately uses when it transmits the message
4442 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4445 50. If a filter generated a file delivery with a non-absolute name (possible if
4446 no home directory exists for the router), the forbid_file option was not
4449 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4450 the existing equality matching.
4452 52. Exim was using ints instead of ino_t variables in some places where it was
4453 dealing with inode numbers.
4455 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4456 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4457 called TMPDIR, and if it finds it is different, it changes its value.
4459 54. The smtp_printf() function is now made available to local_scan() so
4460 additional output lines can be written before returning. There is also an
4461 smtp_fflush() function to enable the detection of a dropped connection.
4462 The variables smtp_input and smtp_batched_input are exported to
4465 55. Changed the default runtime configuration: the message "Unknown user"
4466 has been removed from the ACL, and instead placed on the localuser router,
4467 using the cannot_route_message feature. This means that any verification
4468 failures that generate their own messages won't get overridden. Similarly,
4469 the "Unrouteable address" message that was in the ACL for unverifiable
4470 relay addresses has also been removed.
4472 56. Added hosts_avoid_esmtp to the smtp transport.
4474 57. The exicyclog script was not checking for the esoteric option
4475 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4476 will work only if exicyclog is run under the appropriate euid.
4478 58. Following a discussion on the list, the rules by which Exim recognises line
4479 endings on incoming messages have been changed. The -dropcr and drop_cr
4480 options are now no-ops, retained only for backwards compatibility. The
4481 following line terminators are recognized: LF CRLF CR. However, special
4482 processing applies to CR:
4484 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4485 nor a local message in the state where . is a terminator.
4487 (ii) If a bare CR is encountered in a header line, an extra space is added
4488 after the line terminator so as not to end the header. The reasoning
4489 behind this is that bare CRs in header lines are most likely either
4490 to be mistakes, or people trying to play silly games.
4492 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4493 was being incorrectly given as 18 bytes larger than it should have been.
4494 This is a VOB (very old bug).
4496 60. This may never have affected anything current, but just in case it has:
4497 When the local host is found other than at the start of a list of hosts,
4498 the local host, those with the same MX, and any that follow, are discarded.
4499 When the list in question was part of a longer list of hosts, the following
4500 hosts (not currently being processed) were also being discarded. This no
4501 longer happens. I'm not sure if this situation could ever has previously
4504 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4507 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4508 header. This is something that is recommended in a new Internet Draft, and
4509 is something that is documented as being done by Sendmail. There are two
4510 possible values. For messages generated by the autoreply transport, Exim
4513 Auto-Submitted: auto-replied
4515 whereas for all other generated messages (e.g. bounces) it adds
4517 Auto-Submitted: auto-generated
4519 63. The "personal" condition in filters now includes a test for the
4520 Auto-Submitted: header. If it contains the string "auto-" the message it
4521 not considered personal.
4523 64. Added rcpt_include_affixes as a generic transport option.
4525 65. Added queue_only_override (default true).
4527 66. Added the syslog_duplication option.
4529 67. If what should have been the first header line of a message consisted of
4530 a space followed by a colon, Exim was mis-interpreting it as a header line.
4531 It isn't of course - it is syntactically invalid and should therefore be
4532 treated as the start of the message body. The misbehaviour could have
4533 caused a number of strange effects, including loss of data in subsequent
4534 header lines, and spool format errors.
4536 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4537 client host had authenticated. This control can now be exercised by an ACL
4538 for more flexibility.
4540 69. By default, callouts do not happen when testing with -bh. There is now a
4541 variant, -bhc, which does actually run the callout code, including
4542 consulting and updating the callout cache.
4544 70. Added support for saslauthd authentication, courtesy of Alexander
4547 71. If statvfs() failed on the spool or log directories while checking their
4548 size for availability, Exim confusingly gave the error "space shortage".
4549 Furthermore, in debugging mode it crashed with a floating point exception.
4550 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4551 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4552 serious problem, Exim now writes to the main and panic logs when this
4553 happens, with details of the failure. It then refuses to accept the
4554 incoming message, giving the message "spool directory problem" or "log
4555 directory problem" with a 421 code for SMTP messages.
4557 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4558 0, 1, and 2 exist, because some OS complain for execs without them (see
4559 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4560 descriptors. However, the code omitted to check that the open succeeded,
4561 causing mysterious errors if for some reason the permissions on /dev/null
4562 got screwed. Now Exim writes a message to the main and panic logs, and
4563 bombs out if it can't open /dev/null.
4565 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4566 interact so that it is all more flexible. It is supposed to remain
4567 backwards compatible. Also added extra_local_interfaces.
4569 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4570 to bomb out with an assertion failure - to the client this appears as a
4571 connection drop. This problem occurs in the part of the code that was taken
4572 from the Samba project. Fortunately, the assertion is in a very simple
4573 function, so I have fixed this by reproducing the function inline in the
4574 one place where it is called, and arranging for authentication to fail
4575 instead of killing the process with assert().
4577 75. The SPA client code was not working when the server requested OEM rather
4578 than Unicode encoding.
4580 76. Added code to make require_files with a specific uid setting more usable in
4581 the case where statting the file as root fails - usually a non-root-mounted
4582 NFS file system. When this happens and the failure is EACCES, Exim now
4583 forks a subprocess and does the per-uid checking as the relevant uid.
4585 77. Added process_log_path.
4587 78. If log_file_path was not explicitly set, a setting of check_log_space or
4588 check_log_inodes was ignored.
4590 79. If a space check for the spool or log partitions fails, the incident is now
4591 logged. Of course, in the latter case the data may get lost...
4593 80. Added the %p formatting code to string_format() so that it can be used to
4594 print addresses in debug_print(). Adjusted all the address printing in the
4595 debugging in store.c to use %p rather than %d.
4597 81. There was a concern that a line of code in smtp_in.c could overflow a
4598 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4599 initially expressed, the concern was not well-founded, because trailing
4600 spaces are removed early. However, if the trailing spaces were followed by
4601 a NULL, they did not get removed, so the overflow was possible. Two fixes
4604 (a) I re-wrote the offending code in a cleaner fashion.
4605 (b) If an incoming SMTP command contains a NULL character, it is rejected
4608 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4609 runs initgroups(), so that if the Exim user is in any additional groups,
4610 they will be used during message reception.
4616 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob