1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.464 2007/01/30 11:45:20 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.
69 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
70 fixed by 4.65/MH/01 (is this a record?) are fixed:
72 (i) An empty string was always treated as zero by the numeric comparison
73 operators. This behaviour has been restored.
75 (ii) It is documented that the numeric comparison operators always treat
76 their arguments as decimal numbers. This was broken in that numbers
77 starting with 0 were being interpreted as octal.
79 While fixing these problems I realized that there was another issue that
80 hadn't been noticed. Values of message_size_limit (both the global option
81 and the transport option) were treated as octal if they started with 0.
82 The documentation was vague. These values are now always treated as
83 decimal, and I will make that clear in the documentation.
89 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
90 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
93 MH/01 Don't check that the operands of numeric comparison operators are
94 integers when their expansion is in "skipping" mode (fixes bug
95 introduced by 4.64-PH/07).
97 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
98 child addresses, Exim now panics and dies. Previously, because the count
99 is held in a short int, deliveries were likely to be lost. As such a
100 large number of recipients for a single message is ridiculous
101 (performance will be very, very poor), I have chosen to impose a limit
102 rather than extend the field.
108 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
109 leftover -K file (the existence of which was triggered by #402).
110 While we were at it, introduced process PID as part of the -K
111 filename. This should rule out race conditions when creating
114 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
115 processing considerably. Previous code took too long for large mails,
116 triggering a timeout which in turn triggers #401.
118 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
119 in the DK code in transports.c. sendfile() is not really portable,
120 hence the _LINUX specificness.
122 TF/01 In the add_headers option to the mail command in an Exim filter,
123 there was a bug that Exim would claim a syntax error in any
124 header after the first one which had an odd number of characters
127 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
128 callout verification, Exim cached a "reject" for the entire domain. This
129 is correct for most verifications, but it is not correct for a recipient
130 verification with use_sender or use_postmaster set, because in that case
131 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
132 case of MAIL FROM:<> rejection from other early rejections (e.g.
133 rejection of HELO). When verifying a recipient using a non-null MAIL
134 address, the cache is ignored if it shows MAIL FROM:<> rejection.
135 Whatever the result of the callout, the value of the domain cache is
136 left unchanged (for any other kind of callout, getting as far as trying
137 RCPT means that the domain itself is ok).
139 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
142 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
143 manifest itself as EPIPE rather than ECONNECT. When tidying away a
144 session, the daemon ignores ECONNECT errors and logs others; it now
145 ignores EPIPE as well.
147 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
148 (quoted-printable decoding).
150 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
151 later the small subsequent patch to fix an introduced bug.
153 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
155 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
157 PH/08 An error is now given if message_size_limit is specified negative.
159 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
160 to be given (somewhat) arbitrary names.
162 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
165 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
166 miscellaneous code fixes
168 PH/10 Added the log_reject_target ACL modifier to specify where to log
171 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
172 hostname. This is wrong, because it relates to the incoming message (and
173 probably the interface on which it is arriving) and not to the outgoing
174 callout (which could be using a different interface). This has been
175 changed to use the value of the helo_data option from the smtp transport
176 instead - this is what is used when a message is actually being sent. If
177 there is no remote transport (possible with a router that sets up host
178 addresses), $smtp_active_hostname is used.
180 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
181 tweaks were necessary in order to get it to work (see also 21 below):
182 (a) The code assumed that strncpy() returns a negative number on buffer
183 overflow, which isn't the case. Replaced with Exim's string_format()
185 (b) There were several signed/unsigned issues. I just did the minimum
186 hacking in of casts. There is scope for a larger refactoring.
187 (c) The code used strcasecmp() which is not a standard C function.
188 Replaced with Exim's strcmpic() function.
189 (d) The code set only $1; it now sets $auth1 as well.
190 (e) A simple test gave the error "authentication client didn't specify
191 service in request". It would seem that Dovecot has changed its
192 interface. Fortunately there's a specification; I followed it and
193 changed what the client sends and it appears to be working now.
195 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
198 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
199 address A is aliased to B and C, where B exists and C does not. Without
200 -v the output is "A verified" because verification stops after a
201 successful redirection if more than one address is generated. However,
202 with -v the child addresses are also verified. Exim was outputting "A
203 failed to verify" and then showing the successful verification for C,
204 with its parentage. It now outputs "B failed to verify", showing B's
205 parentage before showing the successful verification of C.
207 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
208 look up a TXT record in a specific list after matching in a combined
211 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
212 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
213 they consult the DNS. I had assumed they would set it the way they
214 wanted; and indeed my experiments on Linux seem to show that in some
215 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
216 To be on the safe side, however, I have now made the interface to
217 host_find_byname() similar to host_find_bydns(), with an argument
218 containing the DNS resolver options. The host_find_byname() function now
219 sets these options at its start, just as host_find_bydns() does. The smtp
220 transport options dns_qualify_single and dns_search_parents are passed to
221 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
222 of host_find_byname() use the default settings of RES_DEFNAMES
223 (qualify_single) but not RES_DNSRCH (search_parents).
225 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
226 spool_read_header() do less string testing, by means of a preliminary
227 switch on the second character of optional "-foo" lines. (This is
228 overdue, caused by the large number of possibilities that now exist.
229 Originally there were few.) While I was there, I also converted the
230 str(n)cmp tests so they don't re-test the leading "-" and the first
231 character, in the hope this might squeeze out yet more improvement.
233 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
234 flag allowing group syntax was set by the header_syntax check but not
235 turned off, possible causing trouble later; (2) The flag was not being
236 set at all for the header_verify test, causing "group"-style headers to
237 be rejected. I have now set it in this case, and also caused header_
238 verify to ignore an empty address taken from a group. While doing this, I
239 came across some other cases where the code for allowing group syntax
240 while scanning a header line wasn't quite right (mostly, not resetting
241 the flag correctly in the right place). These bugs could have caused
242 trouble for malformed header lines. I hope it is now all correct.
244 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
245 with the "reply" argument non-NULL. The code, however (which originally
246 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
247 but it didn't always do it. This confused somebody who was copying the
248 code for some other use. I have removed all the tests.
250 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
251 feature that was used to support insecure browsers during the U.S. crypto
252 embargo. It requires special client support, and Exim is probably the
253 only MTA that supported it -- and would never use it because real RSA is
254 always available. This code has been removed, because it had the bad
255 effect of slowing Exim down by computing (never used) parameters for the
256 RSA_EXPORT functionality.
258 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
259 authenticator to fail if there's a tab character in the incoming data
260 (there should never be unless someone is messing about, as it's supposed
261 to be base64-encoded). Also added, on Timo's advice, the "secured" option
262 if the connection is using TLS or if the remote IP is the same as the
263 local IP, and the "valid-client-cert option" if a client certificate has
266 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
267 authenticators. This can be used for authorization after authentication
268 succeeds. (In the case of plaintext, it servers for both authentication
271 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
272 if any retry times were supplied.
274 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
275 connection, where there is no client IP address to check. In this
276 situation, the verify now always succeeds.
278 PH/25 Applied John Jetmore's -Mset patch.
280 PH/26 Added -bem to be like -Mset, but loading a message from a file.
282 PH/27 In a string expansion for a processed (not raw) header when multiple
283 headers of the same name were present, leading whitespace was being
284 removed from all of them, but trailing whitespace was being removed only
285 from the last one. Now trailing whitespace is removed from each header
286 before concatenation. Completely empty headers in a concatenation (as
289 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
290 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
292 PH/29 [Removed. This was a change that I later backed out, and forgot to
293 correct the ChangeLog entry (that I had efficiently created) before
294 committing the later change.]
296 PH/30 Exim was sometimes attempting to deliver messages that had suffered
297 address errors (4xx response to RCPT) over the same connection as other
298 messages routed to the same hosts. Such deliveries are always "forced",
299 so retry times are not inspected. This resulted in far too many retries
300 for the affected addresses. The effect occurred only when there were more
301 hosts than the hosts_max_try setting in the smtp transport when it had
302 the 4xx errors. Those hosts that it had tried were not added to the list
303 of hosts for which the message was waiting, so if all were tried, there
304 was no problem. Two fixes have been applied:
306 (i) If there are any address or message errors in an SMTP delivery, none
307 of the hosts (tried or untried) are now added to the list of hosts
308 for which the message is waiting, so the message should not be a
309 candidate for sending over the same connection that was used for a
310 successful delivery of some other message. This seems entirely
311 reasonable: after all the message is NOT "waiting for some host".
312 This is so "obvious" that I'm not sure why it wasn't done
313 previously. Hope I haven't missed anything, but it can't do any
314 harm, as the worst effect is to miss an optimization.
316 (ii) If, despite (i), such a delivery is accidentally attempted, the
317 routing retry time is respected, so at least it doesn't keep
318 hammering the server.
320 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
321 in ${readsocket because some servers need this prod.
323 PH/32 Added some extra debug output when updating a wait-xxx database.
325 PH/33 The hint "could be header name not terminated by colon", which has been
326 given for certain expansion errors for a long time, was not being given
327 for the ${if def:h_colon_omitted{... case.
329 PH/34 The spec says: "With one important exception, whenever a domain list is
330 being scanned, $domain contains the subject domain." There was at least
331 one case where this was not true.
333 PH/35 The error "getsockname() failed: connection reset by peer" was being
334 written to the panic log as well as the main log, but it isn't really
335 panic-worthy as it just means the connection died rather early on. I have
336 removed the panic log writing for the ECONNRESET error when getsockname()
339 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
340 runs only) independently of the message's sender address. This meant
341 that, if the 4xx error was in fact related to the sender, a different
342 message to the same recipient with a different sender could confuse
343 things. In particualar, this can happen when sending to a greylisting
344 server, but other circumstances could also provoke similar problems.
345 I have changed the default so that the retry time for these errors is now
346 based a combination of the sender and recipient addresses. This change
347 can be overridden by setting address_retry_include_sender=false in the
350 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
351 remote server are returned as part of bounce messages. This was not
352 happening for LMTP over a pipe (the lmtp transport), but now it is the
353 same for both kinds of LMTP.
355 PH/38 Despite being documented as not happening, Exim was rewriting addresses
356 in header lines that were in fact CNAMEs. This is no longer the case.
358 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
359 and queue runs started by the daemon processed all messages. This has
360 been fixed so that -R and -S can now usefully be given with -q<time>.
362 PH/40 Import PCRE release 6.7 (fixes some bugs).
364 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
366 PH/42 Give an error if -q is specified more than once.
368 PH/43 Renamed the variables $interface_address and $interface_port as
369 $received_ip_address and $received_port, to make it clear that these
370 values apply to message reception, and not to the outgoing interface when
371 a message is delivered. (The old names remain recognized, of course.)
373 PH/44 There was no timeout on the connect() call when using a Unix domain
374 socket in the ${readsocket expansion. There now is.
376 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
377 be meaningful with "accept".
379 SC/01 Eximstats V1.43
380 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
382 SC/02 Eximstats V1.44
383 Use a glob alias rather than an array ref in the generated
384 parser. This improves both readability and performance.
386 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
387 Collect SpamAssassin and rejection statistics.
388 Don't display local sender or destination tables unless
389 there is data to show.
390 Added average volumes into the top table text output.
392 SC/04 Eximstats V1.46
393 Collect data on the number of addresses (recipients)
394 as well as the number of messages.
396 SC/05 Eximstats V1.47
397 Added 'Message too big' to the list of mail rejection
398 reasons (thanks to Marco Gaiarin).
400 SC/06 Eximstats V1.48
401 Mainlog lines which have GMT offsets and are too short to
402 have a flag are now skipped.
404 SC/07 Eximstats V1.49 (Alain Williams)
405 Added the -emptyok flag.
407 SC/08 Eximstats V1.50
408 Fixes for obtaining the IP address from reject messages.
410 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
411 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
412 whitesspace changes from 4.64-PH/27
414 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
417 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
418 are found, allow negative numbers in numeric criteria)
420 JJ/06 exipick.20061117.2, added new $message_body_missing variable
422 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
423 to match changes made in 4.64-PH/43
425 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
427 PH/47 Put in an explicit test for a DNS lookup of an address record where the
428 "domain" is actually an IP address, and force a failure. This locks out
429 those revolvers/nameservers that support "A-for-A" lookups, in
430 contravention of the specifications.
432 PH/48 When a host name was looked up from an IP address, and the subsequent
433 forward lookup of the name timed out, the host name was left in
434 $sender_host_name, contrary to the specification.
436 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
437 restricted to single-key lookups, Exim was not diagnosing an error if
438 * or *@ was used with a query-style lookup.
440 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
442 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
443 long ago, but noone interested enough thought of it. Let's just say that
444 the "1.1" means that there are some new functions that weren't there at
445 some point in the past.
447 PH/51 Error processing for expansion failure of helo_data from an smtp
448 transport during callout processing was broken.
450 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
451 tested/used via the -bh/-bhc/-bs options.
453 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
454 bug, fixed in subsequent PCRE releases).
456 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
457 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
459 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
465 SC/01 Use a glob alias rather than an array ref in eximstats generated
466 parser. This improves both readability and performance.
468 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
469 Don't display local sender or destination tables in eximstats unless
470 there is data to show.
471 Added average volumes into the eximstats top table text output.
473 SC/03 Collect data on the number of addresses (recipients) as well
474 as the number of messages in eximstats.
476 TF/01 Correct an error in the documentation for the redirect router. Exim
477 does (usually) call initgroups() when daemonizing.
479 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
480 with consistent privilege compared to when running as a daemon.
482 TF/03 Note in the spec that $authenticated_id is not set for local
483 submissions from trusted users.
485 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
486 Thanks to Dean Brooks <dean@iglou.com> for the patch.
488 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
489 by adding some example configuration directives to the default
490 configuration file. A little bit of work is required to uncomment the
491 directives and define how usernames and passwords are checked, but
492 there is now a framework to start from.
494 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
495 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
496 without this. I don't know how relevant this is to other LDAP libraries.
498 PH/02 Add the verb name to the "unknown ACL verb" error.
500 PH/03 Magnus Holmgren's patch for filter_prepend_home.
502 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
504 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
505 directory not expanded when it should be if an expanded home directory
506 was set for the address (which is overridden by the transport).
508 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
511 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
512 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
513 because it is too late at that time, and has no effect.
515 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
516 security issue with \' (bugzilla #107). I could not use the
517 PQescapeStringConn() function, because it needs a PGconn value as one of
520 PH/08 When testing addresses using -bt, indicate those final addresses that
521 are duplicates that would not cause an additional delivery. At least one
522 person was confused, thinking that -bt output corresponded to deliveries.
523 (Suppressing duplicates isn't a good idea as you lose the information
524 about possibly different redirections that led to the duplicates.)
526 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
527 systems where poll() doesn't work, in particular OS X.
529 PH/10 Added more information to debugging output for retry time not reached.
531 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
532 operations in malware.c.
534 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
537 PH/13 If write_rejectlog was set false when logging was sent to syslog with
538 syslog_duplication set false, log lines that would normally be written
539 both the the main log and to the reject log were not written to syslog at
542 PH/14 In the default configuration, change the use of "message" in ACL warn
543 statements to "add_header".
545 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
546 not followed by a command (e.g. "seen endif").
548 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
549 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
552 PH/17 Added extra conditions to the default value of delay_warning_condition
556 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
557 { match{$h_precedence:}{(?i)bulk|list|junk} } \
558 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
561 The Auto-Submitted: and various List- headers are standardised, whereas I
562 don't think Precedence: ever was.
564 PH/18 Refactored debugging code in route_finduser() to show more information,
565 in particular, the error code if getpwnam() issues one.
567 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
568 This is apparently needed in addition to the PH/07 change above to avoid
569 any possible encoding problems.
571 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
572 but not after initializing Perl.
574 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
575 output them only if debugging. By default they are written stderr,
576 apparently, which is not desirable.
578 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
581 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
584 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
586 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
587 authenticated or an ident call has been made. Suppress the default
588 values for $authenticated_id and $authenticated_sender (but permit -oMai
589 and -oMas) when testing with -bh.
591 PH/24 Re-jigged the order of the tests in the default configuration so that the
592 tests for valid domains and recipients precede the DNS black list and CSA
593 tests, on the grounds that those ones are more expensive.
595 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
596 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
597 This bug exists in every version of Exim that I still have, right back to
600 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
601 However, an attempt to turn on case-sensitivity in a regex key by
602 including (?-i) didn't work because the subject string was already
603 lowercased, and the effects were non-intuitive. It turns out that a
604 one-line patch can be used to allow (?-i) to work as expected.
610 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
611 other effects) broke the use of negated acl sub-conditions.
613 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
616 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
617 "Deny" causes Exim to reject the incoming connection with a 554 error.
618 Unfortunately, if there is a major crisis, such as a disk failure,
619 tcp-wrappers gives "deny", whereas what one would like would be some
620 kind of temporary error. A kludge has been added to help with this.
621 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
622 554 error is used if errno is still zero or contains ENOENT (which occurs
623 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
626 PH/03 Add -lutil to the default FreeBSD LIBS setting.
628 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
629 errors. Otherwise a message that provokes a temporary error (when other
630 messages do not) can cause a whole host to time out.
632 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
633 addresses were routed directly to files or pipes from a redirect router.
634 File deliveries just didn't batch; pipe deliveries might have suffered
637 PH/06 A failure to get a lock for a hints database would erroneously always say
638 "Failed to get write lock", even when it was really a read lock.
640 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
641 of 0600. This has been changed to use the value of the lockfile_mode
642 option (which defaults to 0600).
644 PH/08 Applied small patch from the Sieve maintainer.
646 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
647 folder from quota calculations, a direct delivery into this folder messed
648 up the contents of the maildirsize file. This was because the regex was
649 used only to exclude .Trash (or whatever) when the size of the mailbox
650 was calculated. There was no check that a delivery was happening into an
651 excluded directory. This bug has been fixed by ignoring all quota
652 processing for deliveries into excluded directories.
654 PH/10 Added the maildirfolder_create_regex option to appendfile.
660 PH/01 The code for finding all the local interface addresses on a FreeBSD
661 system running IPv6 was broken. This may well have applied to all BSD
662 systems, as well as to others that have similar system calls. The broken
663 code found IPv4 interfaces correctly, but gave incorrect values for the
664 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
665 that it would not match correctly against @[] and not recognize the IPv6
668 PH/02 The ipliteral router was not recognizing addresses of the form user@
669 [ipv6:....] because it didn't know about the "ipv6:" prefix.
671 PH/03 Added disable_ipv6.
673 PH/04 Changed $reply_address to use the raw form of the headers instead of the
674 decoded form, because it is most often used to construct To: headers
675 lines in autoreplies, and the decoded form may well be syntactically
676 invalid. However, $reply_address has leading white space removed, and all
677 newlines turned into spaces so that the autoreply transport does not
680 PH/05 If group was specified without a user on a router, and no group or user
681 was specified on a transport, the group from the router was ignored.
683 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
684 for visible compile-time settings that can be used to change these
685 numbers, for those that want even more. Backwards compatibility with old
686 spool files has been maintained. However, going back to a previous Exim
687 release will lost any variables that are in spool files.
689 PH/07 Two small changes when running in the test harness: increase delay when
690 passing a TCP/IP connection to a new process, in case the original
691 process has to generate a bounce, and remove special handling of
692 127.0.0.2 (sic), which is no longer necessary.
694 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
695 be the same on different OS.
697 PH/09 Moved a debug statement in filter processing to avoid a race problem when
700 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
701 whether --show-vars was specified or not
703 JJ/02 exipick: Added support for new ACL variable spool format introduced
706 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
707 syntactically invalid From: or Reply-to: line, and a filter used this to
708 generate an autoreply, and therefore failed to obtain an address for the
709 autoreply, Exim could try to deliver to a non-existent relative file
710 name, causing unrelated and misleading errors. What now happens is that
711 it logs this as a hard delivery error, but does not attempt to create a
714 PH/11 The exinext utility has a -C option for testing purposes, but although
715 the given file was scanned by exinext itself; it wasn't being passed on
716 when Exim was called.
718 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
719 an end-of-file indication when reading a command response.
721 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
722 compiled. In many other places in Exim, IPv6 addresses are always
723 recognized, so I have changed this. It also means that IPv4 domain
724 literals of the form [IPV4:n.n.n.n] are now always recognized.
726 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
727 used if the router is not running as root, for example, when verifying at
728 ACL time, or when using -bh. The debugging output from this situation was
729 non-existent - all you got was a failure to exec. I have made two
732 (a) Failures to set uid/gid, the current directory, or a process leader
733 in a subprocess such as that created by queryprogram now generate
734 suitable debugging ouput when -d is set.
736 (b) The queryprogram router detects when it is not running as root,
737 outputs suitable debugging information if -d is set, and then runs
738 the subprocess without attempting to change uid/gid.
740 PH/15 Minor change to Makefile for building test_host (undocumented testing
743 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
744 additional section of a DNS packet that returns MX or SRV records.
745 Instead, it always explicitly searches for A/AAAA records. This avoids
746 major problems that occur when a DNS server includes only records of one
747 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
748 fixed another bug: if SRV records were looked up and the corresponding
749 address records were *not* found in the additional section, the port
750 values from the SRV records were lost.
752 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
753 using the correct key (the original address) when searching the retry
754 rules in order to find which one to use for generating the retry hint.
756 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
757 adding the default one. Similarly, if it contains a Reply-To: header, the
758 errors_reply_to option, if set, is not used.
760 PH/19 When calculating a retry time, Exim used to measure the "time since
761 failure" by looking at the "first failed" field in the retry record. Now
762 it does not use this if it is later than than the arrival time of the
763 message. Instead it uses the arrival time. This makes for better
764 behaviour in cases where some deliveries succeed, thus re-setting the
765 "first failed" field. An example is a quota failure for a huge message
766 when small messages continue to be delivered. Without this change, the
767 "time since failure" will always be short, possible causing more frequent
768 delivery attempts for the huge message than are intended.
769 [Note: This change was subsequently modified - see PH/04 for 4.62.]
771 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
772 $1, $2, $3) because the numerical variables can be reset during some
773 expansion items (e.g. "match"), thereby losing the authentication data.
775 PH/21 Make -bV show the size of off_t variables so that the test suite can
776 decide whether to run tests for quotas > 2G.
778 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
779 mailbox_size, and mailbox_filecount in the appendfile transport. If a
780 filecount value is greater than 2G or if a quota value is greater than 2G
781 on a system where the size of off_t is not greater than 4, a panic error
784 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
785 never match. The debug and -bh output now contains an explicit error
786 message indicating a malformed IPv4 address or mask.
788 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
789 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
792 PH/25 Do not write to syslog when running in the test harness. The only
793 occasion when this arises is a failure to open the main or panic logs
794 (for which there is an explicit test).
796 PH/26 Added the /no_tell option to "control=freeze".
798 PH/27 If a host name lookup failed very early in a connection, for example, if
799 the IP address matched host_lookup and the reverse lookup yielded a name
800 that did not have a forward lookup, an error message of the form "no IP
801 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
802 could be logged. Now it outputs the IP address instead of "NULL".
804 PH/28 An enabling patch from MH: add new function child_open_exim2() which
805 allows the sender and the authenticated sender to be set when
806 submitting a message from within Exim. Since child_open_exim() is
807 documented for local_scan(), the new function should be too.
809 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
810 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
811 results in an empty string is now treated as unset.
813 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
815 PH/31 Added sender_verify_fail logging option.
817 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
818 needed by RFC 822 but not by RFC 2822 was commented out. I have now
819 tidied the source and removed it altogether.
821 PH/33 When a queue run was abandoned because the load average was too high, a
822 log line was always written; now it is written only if the queue_run log
823 selector is set. In addition, the log line for abandonment now contains
824 information about the queue run such as the pid. This is always present
825 in "start" and "stop" lines but was omitted from the "abandon" line.
827 PH/34 Omit spaces between a header name and the colon in the error message that
828 is given when verify = headers_syntax fails (if there are lots of them,
829 the message gets confusing).
831 PH/35 Change the default for dns_check_names_pattern to allow slashes within
832 names, as there are now some PTR records that contain slashes. This check
833 is only to protect against broken name servers that fall over on strange
834 characters, so the fact that it applies to all lookups doesn't matter.
836 PH/36 Now that the new test suite is complete, we can remove some of the
837 special code in Exim that was needed for the old test suite. For example,
838 sorting DNS records because real resolvers return them in an arbitrary
839 order. The new test suite's fake resolver always returns records in the
842 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
843 bounces) except when queue_only is set, to avoid logging races between
844 the different processes.
846 PH/38 Panic-die if .include specifies a non-absolute path.
848 PH/39 A tweak to the "H" retry rule from its user.
850 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
851 a label. They prevented compilation on older perls.
853 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
854 a warning to be raised on newish perls.
856 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
857 on queue. Changes to match documented behaviour of showing count of
858 messages matching specified criteria.
860 PH/40 Changed the default ident timeout from 30s to 5s.
862 PH/41 Added support for the use of login_cap features, on those BSD systems
863 that have them, for controlling the resources used by pipe deliveries.
865 PH/42 The content-scanning code uses fopen() to create files in which to put
866 message data. Previously it was not paying any attention to the mode of
867 the files. Exim runs with umask(0) because the rest of the code creates
868 files with open(), and sets the required mode explicitly. Thus, these
869 files were ending up world-writeable. This was not a big issue, because,
870 being within the spool directory, they were not world-accessible. I have
871 created a function called modefopen, which takes an additional mode
872 argument. It sets umask(777), creates the file, chmods it to the required
873 mode, then resets the umask. All the relevant calls to fopen() in the
874 content scanning code have been changed to use this function.
876 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
877 to 24 hours. This avoids potential overflow problems when processing G
878 and H retry rules. I suspect nobody ever tinkers with this value.
880 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
882 PH/45 When the plaintext authenticator is running as a client, the server's
883 challenges are checked to ensure they are valid base64 strings. By
884 default, the authentication attempt is cancelled if an invalid string is
885 received. Setting client_ignore_invalid_base64 true ignores these errors.
886 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
887 they are received. Thus, the responses can be made to depend on the
888 challenges. If an invalid string is ignored, an empty string is placed in
891 PH/46 Messages that are created by the autoreply transport now contains a
892 References: header, in accordance with RFCs 2822 and 3834.
894 PH/47 Added authenticated_sender_force to the smtp transport.
896 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
898 PH/49 Installed latest patch from the Sieve maintainer.
900 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
901 was also set, the appendfile transport was unnecessarily scanning a
902 directory of message files (e.g. for maildir delivery) to find the count
903 of files (along with the size), even though it did not need this
904 information. It now does the scan only if it needs to find either the
905 size of the count of files.
907 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
909 PH/52 Two bugs concerned with error handling when the smtp transport is
912 (i) Exim was not creating retry information for temporary errors given
913 for individual recipients after the DATA command when the smtp transport
914 was used in LMTP mode. This meant that they could be retried too
915 frequently, and not timed out correctly.
917 (ii) Exim was setting the flag that allows error details to be returned
918 for LMTP errors on RCPT commands, but not for LMTP errors for individual
919 recipients that were returned after the DATA command.
921 PH/53 This is related to PH/52, but is more general: for any failing address,
922 when detailed error information was permitted to be returned to the
923 sender, but the error was temporary, then after the final timeout, only
924 "retry timeout exceeded" was returned. Now it returns the full error as
925 well as "retry timeout exceeded".
927 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
928 do this, and (what is worse) MTAs that accept it.
930 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
931 will now be deprecated.
933 PH/56 New os.c-cygwin from the Cygwin maintainer.
935 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
936 formats (previously only available in exim formats via -bpr, -bpru,
937 and -bpra. Now also available in native and exiqgrep formats)
939 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
940 with very large, slow to parse queues
942 JJ/08 exipick: added ! as generic prefix to negate any criteria format
944 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
946 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
947 responses to authentication challenges, though it was showing the
948 challenges; (ii) I've removed the CR characters from the debug output for
951 PH/58 Allow for the insertion of a newline as well as a space when a string
952 is turned into more than one encoded-word during RFC 2047 encoding. The
953 Sieve code now uses this.
955 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
956 data_4xx, lost_connection, tls_required.
958 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
959 message was being sent as an SMTP response.
961 PH/61 Add -l and -k options to exicyclog.
963 PH/62 When verifying, if an address was redirected to one new address, so that
964 verification continued, and the new address failed or deferred after
965 having set something in $address_data, the value of $address_data was not
966 passed back to the ACL. This was different to the case when no
967 redirection occurred. The value is now passed back in both cases.
969 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
970 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
971 use login_cap.h, so on its own it isn't the distinguishing feature. The
972 new name refers directly to the setclassresources() function.
974 PH/65 Added configuration files for NetBSD3.
976 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
978 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
979 is preferred over IPv4.
981 PH/68 The bounce_return_message and bounce_return_body options were not being
982 honoured for bounces generated during the reception of non-SMTP messages.
983 In particular, this applied to messages rejected by the ACL. This bug has
984 been fixed. However, if bounce_return_message is true and bounce_return_
985 body is false, the headers that are returned for a non-SMTP message
986 include only those that have been read before the error was detected.
987 (In the case of an ACL rejection, they have all been read.)
989 PH/69 The HTML version of the specification is now built in a directory called
990 spec_html instead of spec.html, because the latter looks like a path with
991 a MIME-type, and this confuses some software.
993 PH/70 Catch two compiler warnings in sieve.c.
995 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
996 function verify_get_ident() calls ip_connect() to connect a socket, but
997 if the "connect()" function timed out, ip_connect() used to close the
998 socket. However, verify_get_ident() also closes the socket later, and in
999 between Exim writes to the log, which may get opened at this point. When
1000 the socket was closed in ip_connect(), the log could get the same file
1001 descriptor number as the socket. This naturally causes chaos. The fix is
1002 not to close the socket in ip_connect(); the socket should be closed by
1003 the function that creates it. There was only one place in the code where
1004 this was missing, in the iplookup router, which I don't think anybody now
1005 uses, but I've fixed it anyway.
1007 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1008 well as to direct DNS lookups. Otherwise the handling of names in host
1009 lists is inconsistent and therefore confusing.
1015 PH/01 Two changes to the default runtime configuration:
1017 (1) Move the checks for relay_from_hosts and authenticated clients from
1018 after to before the (commented out) DNS black list checks.
1020 (2) Add control=submission to the relay_from_hosts and authenticated
1021 clients checks, on the grounds that messages accepted by these
1022 statements are most likely to be submissions.
1024 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1026 (1) Generate an error if the third argument for the ${prvs expansion is
1029 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1032 (3) Reset the variables that are obtained from the first argument of
1033 ${prvscheck and used in the second argument before leaving the code,
1034 because their memory is reclaimed, so using them afterwards may do
1037 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1038 one (it's much easier than Tom thought :-).
1040 (5) Because of (4), we can now allow for the use of $prvscheck_result
1041 inside the third argument.
1043 PH/03 For some reason, the default setting of PATH when running a command from
1044 a pipe transport was just "/usr/bin". I have changed it to
1047 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1048 anything to be listed in the output from -bV.
1050 PH/05 When a filter generated an autoreply, the entire To: header line was
1051 quoted in the delivery log line, like this:
1053 => >A.N.Other <ano@some.domain> <original@ddress> ...
1055 This has been changed so that it extracts the operative address. There
1056 may be more than one such address. If so, they are comma-separated, like
1059 => >ano@some.domain,ona@other.domain <original@ddress> ...
1061 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1062 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1063 not being looked up in the rDNS to get a host name, Exim was showing the
1064 IP address twice in Received: lines, even though the IP addresses were
1065 identical. For example:
1067 Received: from [1.2.3.4] (helo=[1.2.3.4])
1069 However, if the real host name was known, it was omitting the HELO data
1070 if it matched the actual IP address. This has been tidied up so that it
1071 doesn't show the same IP address twice.
1073 PH/07 When both +timestamp and +memory debugging was on, the value given by
1074 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1075 called by the debug printing, thereby overwriting the timestamp buffer.
1076 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1078 PH/08 When the original message was included in an autoreply transport, it
1079 always said "this is a copy of the message, including all the headers",
1080 even if body_only or headers_only was set. It now gives an appropriate
1083 PH/09 Applied a patch from the Sieve maintainer which:
1085 o fixes some comments
1086 o adds the (disabled) notify extension core
1087 o adds some debug output for the result of if/elsif tests
1088 o points to the current vacation draft in the documentation
1089 and documents the missing references header update
1093 o fixes a bug in processing the envelope test (when testing
1094 multiple envelope elements, the last element determinted the
1097 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1098 Electronic Mail") by including:
1100 Auto-submitted: auto-generated
1102 in the messages that it generates (bounce messages and others, such as
1103 warnings). In the case of bounce messages for non-SMTP mesages, there was
1104 also a typo: it was using "Auto_submitted" (underscore instead of
1105 hyphen). Since every message generated by Exim is necessarily in response
1106 to another message, thes have all been changed to:
1108 Auto-Submitted: auto-replied
1110 in accordance with these statements in the RFC:
1112 The auto-replied keyword:
1114 - SHOULD be used on messages sent in direct response to another
1115 message by an automatic process,
1117 - MUST NOT be used on manually-generated messages,
1119 - MAY be used on Delivery Status Notifications (DSNs) and Message
1120 Disposition Notifications (MDNs),
1122 - MUST NOT be used on messages generated by automatic or periodic
1123 processes, except for messages which are automatic responses to
1126 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1127 to the default Received: header definition.
1129 PH/12 Added log selector acl_warn_skipped (default on).
1131 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1132 variables because (a) they are in the wrong storage pool and (b) even if
1133 they were copied, it wouldn't work properly because of the caching.
1135 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1136 checking when decoding. Apparently there are clients that generate
1137 overlong encoded strings. Why am I not surprised?
1139 PH/15 If the first argument of "${if match_address" was not empty, but did not
1140 contain an "@" character, Exim crashed. Now it writes a panic log message
1141 and treats the condition as false.
1143 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1145 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1146 extension "envelope-auth". The code is finished and in agreement with
1147 other implementations, but there is no documentation so far and in fact,
1148 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1149 not changing the active code.
1151 Print executed "if" and "elsif" statements when debugging is used. This
1152 helps a great deal to understand what a filter does.
1154 Document more things not specified clearly in RFC3028. I had all this
1155 sorted out, when out of a sudden new issues came to my mind. Oops."
1157 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1160 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1161 canonical form (as documented). However, after a host name lookup from
1162 the IP address, check_host() was doing a simple string comparison with
1163 addresses acquired from the DNS when checking that the found name did
1164 have the original IP as one of its addresses. Since any found IPv6
1165 addresses are likely to be in abbreviated form, the comparison could
1166 fail. Luckily, there already exists a function for doing the comparison
1167 by converting both addresses to binary, so now that is used instead of
1168 the text comparison.
1170 PH/20 There was another similar case to PH/19, when a complete host name was
1171 given in a host list; looking up its IP address could give an abbreviated
1172 form, whereas the current host's name might or might not be abbreviated.
1173 The same fix has been applied.
1179 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1180 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1183 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1184 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1186 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1188 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1189 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1190 Also, for "Auto-Submitted", treat anything other than "no" as
1191 non-personal, in accordance with RFC 3834. (Previously it treated
1192 anything starting "auto-" as non-personal.)
1194 TF/01 The control=submission/name=... option had a problem with syntax
1195 errors if the name included a slash character. The /name= option
1196 now slurps the rest of the string, so it can include any characters
1197 but it must come last in the list of options (after /sender_retain
1200 PH/05 Some modifications to the interface to the fake nameserver for the new
1208 TK/01 Added the "success_on_redirect" address verification option. See
1209 NewStuff for rationale and an example.
1211 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1213 PH/02 Patch to exigrep to allow it to work on syslog lines.
1215 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1216 fread() to skip over the body file's header line, because in Cygwin the
1217 header line is locked and is inaccessible.
1219 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1220 co-exist for some time) to make it clear that it is the Exim ID that is
1221 referenced, not the Message-ID: header line.
1223 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1224 string_format() function, because snprintf() does not exist on all
1227 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1228 ${stat: expansion item.
1230 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1231 protocol synchronization error", to keep the pedants happy.
1233 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1234 well as for IRIX systems, when gcc is being used. See the host.c source
1237 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1239 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1242 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1243 transport and to the smtp transport in LMTP mode.
1245 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1247 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1248 run a filter in a subprocess. This could lead to confusion in subsequent
1249 lookups in the parent process. There should also be a search_tidyup() at
1250 the end of the subprocess.
1252 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1253 only if the host matched helo_try_verify_hosts, which caused the
1254 verification to occur when the EHLO/HELO command was issued. The ACL just
1255 tested the remembered result. Now, if a previous verification attempt has
1256 not happened, "verify = helo" does it there and then.
1258 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1260 TK/03 Fix log output including CR from clamd.
1262 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1263 exist provoked a memory error which could cause a segfault.
1265 PH/15 Installed PCRE 6.2
1267 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1269 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1270 of the problem. Specifically, suggested +O2 rather than +O1 for the
1273 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1275 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1276 record, the port was not correctly logged when the outgoing_port log
1277 selector was set (it logged the transort's default port).
1279 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1280 fallback_hosts, and "hosts" in the smtp transport.
1282 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1283 host errors such as "Connection refused".
1285 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1286 authentication with radiusclient 0.4.9:
1288 - Error returned from rc_read_config was caught wrongly
1289 - Username/password not passed on to radius server due to wrong length.
1291 The presumption is that some radiusclient API changes for 4.51/PH/17
1292 were not taken care of correctly. The code is still untested by me (my
1293 Linux distribution still has 0.3.2 of radiusclient), but it was
1294 contributed by a Radius user.
1296 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1297 expanding the "port" option of the smtp transport.
1299 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1300 while reading a MIME header. Thanks to Tom Hughes for a patch.
1302 PH/24 Include config.h inside local_scan.h so that configuration settings are
1305 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1306 that in an ACL for RCPT (for example), you can examine exactly what was
1309 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1310 commands, but it was not correctly comparing the address with the actual
1311 client host address. Thus, it would show the EHLO address in Received:
1312 header lines when this was not necessary.
1314 PH/27 Added the % operator to ${eval:}.
1316 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1317 it should be ignoring failures (because with -C, for example, it has lost
1318 privilege). It wasn't ignoring creation failures other than "already
1321 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1324 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1325 because an input line was too long, either on its own, or by virtue of
1326 too many continuations, the temporary file was not being removed, and the
1327 return code was incorrect.
1329 PH/31 Missing "BOOL" in function definition in filtertest.c.
1331 PH/32 Applied Sieve patches from the maintainer.
1333 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1335 PH/33 Added "verify = not_blind".
1337 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1338 Local/Makefile (with some defaults set). These are used in built scripts
1339 such as exicyclog, but they have never been used in the exim_install
1340 script (though there are many overriding facilities there). I have
1341 arranged that the exim_install script now takes note of these two
1344 PH/35 Installed configuration files for Dragonfly.
1346 PH/36 When a locally submitted message by a trusted user did not contain a
1347 From: header, and the sender address was obtained from -f or from an SMTP
1348 MAIL command, and the trusted user did not use -F to supply a sender
1349 name, $originator_name was incorrectly used when constructing a From:
1350 header. Furthermore, $originator_name was used for submission mode
1351 messages from external hosts without From: headers in a similar way,
1352 which is clearly wrong.
1354 PH/37 Added control=suppress_local_fixups.
1356 PH/38 When log_selector = +received_sender was set, and the addition of the
1357 sender made the log line's construction buffer exactly full, or one byte
1358 less than full, an overflow happened when the terminating "\n" was
1361 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1362 when the result of a list match is failure because a DNS lookup failed.
1364 PH/40 RM_COMMAND is now used in the building process.
1366 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1367 the "build-* directories that it finds.
1369 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1370 address in a domain literal was a prefix of an interface address.
1372 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1373 when verifying a sender address, unless rewrite_headers is false.
1375 PH/44 Wrote a long comment about why errors_to addresses are verified as
1376 recipients, not senders.
1378 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1379 the ratelimit ACL was added.
1381 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1383 PH/46 Added extra information about PostgreSQL errors to the error string.
1385 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1386 avoiding the need to install special zones in a real server. This is
1387 backwards compatible; if it can't find the fake resolver, it drops back.
1388 Thus, both old and new test suites can be run.
1390 TF/02 Added util/ratelimit.pl
1392 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1393 clock is set back in time.
1395 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1396 Candler <B.Candler@pobox.com>.
1398 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1399 OK for addresses that are the result of redirections.
1401 PH/48 A number of further additions for the benefit of the new test suite,
1402 including a fake gethostbyname() that interfaces to the fake DNS resolver
1405 TF/06 The fix for widen_domains has also been applied to qualify_single and
1406 search_parents which are the other dnslookup options that can cause
1409 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1412 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1414 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1415 service" if it encountered an MX record with an empty target hostname.
1416 The message is now "an MX or SRV record indicated no SMTP service".
1418 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1419 if the DNS of the sending site is misconfigured. This is quite a
1420 common situation. This change restores the behaviour of treating a
1421 helo verification defer as a failure.
1423 PH/51 If self=fail was set on a router, the bounce message did not include the
1424 actual error message.
1430 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1432 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1433 command itself ended in error, the underlying message about the transport
1434 filter timeout was being overwritten with the pipe command error. Now the
1435 underlying error message should be appended to the second error message.
1437 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1439 PH/02 Reduce the amount of output that "make" produces by default. Full output
1440 can still be requested.
1442 PH/03 The warning log line about a condition test deferring for a "warn" verb
1443 was being output only once per connection, rather than after each
1444 occurrence (because it was using the same function as for successful
1445 "warn" verbs). This seems wrong, so I have changed it.
1447 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1448 it should not have, which might have caused a crash in the right
1449 circumstances, but probably never did.
1451 PH/04 Installed a modified version of Tony Finch's patch to make submission
1452 mode fix the return path as well as the Sender: header line, and to
1453 add a /name= option so that you can make the user's friendly name appear
1456 TF/03 Added the control = fakedefer ACL modifier.
1458 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1459 Mark Lowes for thorough testing.
1461 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1463 TK/03 Merged latest SRS patch from Miles Wilton.
1465 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1466 with the definition in sysexits.h (which is #included earlier).
1467 Fortunately, Exim does not actually use EX_OK. The code used to try to
1468 preserve the sysexits.h value, by assumimg that macro definitions were
1469 scanned for macro replacements. I have been disabused of this notion,
1470 so now the code just undefines EX_OK before #including unistd.h.
1472 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1473 in the smtp transport. When a block could not be written in a single
1474 write() function, the timeout was being re-applied to each part-write.
1475 This seems wrong - if the receiver was accepting one byte at a time it
1476 would take for ever. The timeout is now adjusted when this happens. It
1477 doesn't have to be particularly precise.
1479 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1480 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1482 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1483 without a domain if the check to <postmaster@domain> fails.
1485 SC/01 Eximstats: added -xls and the ability to specify output files
1486 (patch written by Frank Heydlauf).
1488 SC/02 Eximstats: use FileHandles for outputing results.
1490 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1492 SC/04 Eximstats: fixed display of large numbers with -nvr option
1494 SC/05 Eximstats: fixed merging of reports with empty tables.
1496 SC/06 Eximstats: added the -include_original_destination flag
1498 SC/07 Eximstats: removed tabs and trailing whitespace.
1500 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1502 TK/06 MBOX spool code: Add real "From " MBOX separator line
1503 so the .eml file is really in mbox format (even though
1504 most programs do not really care). Patch from Alex Miller.
1506 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1507 The latter is generated from $received_to and is only set if the
1508 message has one envelope recipient. SA can use these headers,
1509 obviously out-of-the-box. Patch from Alex Miller.
1511 PH/08 The ${def test on a variable was returning false if the variable's
1512 value was "0", contrary to what the specification has always said!
1513 The result should be true unless the variable is empty.
1515 PH/09 The syntax error of a character other than { following "${if
1516 def:variable_name" (after optional whitespace) was not being diagnosed.
1517 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1518 accidental colon was present, for example, could give incorrect results.
1520 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1521 result is used (not including appendfile, where other changes are about
1524 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1525 This involved changing a lot of size variables from int to off_t. It
1526 should work with maildirs and everything.
1528 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1529 spamd dying while we are connected to it.
1531 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1534 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1535 function for BDB 4.3.
1537 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1539 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1540 the sources for PCRE have been re-arranged and more files are now
1543 PH/15 The code I had for printing potentially long long variables in PH/11
1544 above was not the best (it lost precision). The length of off_t variables
1545 is now inspected at build time, and an appropriate printing format (%ld
1546 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1547 to be "long long int" or "long int". This is needed for the internal
1548 formatting function string_vformat().
1550 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1551 the configuration file to be ":syslog", then the script "guesses" where
1552 the logs files are, rather than using the compiled in default. In our
1553 case the guess is not the same as the compiled default, so the script
1554 suddenly stopped working when I started to use syslog. The patch checks
1555 to see if log_file_path is "". If so, it attempts to read it from exim
1556 with no configuration file to get the compiled in version, before it
1557 falls back to the previous guessing code."
1559 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1560 implementing BATV in an Exim configuration. See NewStuff for the gory
1563 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1564 Makefile that are specific to HP-UX.
1566 PH/18 If the "use_postmaster" option was set for a recipient callout together
1567 with the "random" option, the postmaster address was used as the MAIL
1568 FROM address for the random test, but not for the subsequent recipient
1569 test. It is now used for both.
1571 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1572 patch removes a few documentation additions to RFC 3028, because the
1573 latest draft now contains them. It adds the new en;ascii-case comparator
1574 and a new error check for 8bit text in MIME parts. Comparator and
1575 require names are now matched exactly. I enabled the subaddress
1576 extension, but it is not well tested yet (read: it works for me)."
1578 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1579 rework some of the code of TK/09 above to avoid the hardwired use of
1580 "%lld" and "long long". Replaced the call to snprintf() with a call to
1583 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1584 records point to non-existent hosts", "retry timeout exceeded", and
1585 "retry time not reached for any host after a long failure period".
1587 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1588 experimental DomainKeys support:
1590 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1591 (2) On an error such as an illegally used "control", the wrong name for
1592 the control was given.
1594 These problems did NOT occur unless DomainKeys support was compiled.
1596 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1598 PH/24 Added ${if match_ip condition.
1600 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1601 databases so that it will be absolutely obvious if a crash occurs in the
1602 DB library. This is a regular occurrence (often caused by mis-matched
1605 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1606 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1607 picked up on a user's system that detects such things. There doesn't seem
1608 to be a gcc warning option for this - only an attribute that has to be
1609 put on the function's prototype. It seems that in Fedora Core 4 they have
1610 set this on a number of new functions. No doubt there will be more in due
1613 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1614 specify a transport. However, if an address that was verified by such a
1615 router was the subject of a callout, Exim crashed because it tried to
1616 read the rcpt_include_affixes from the non-existent transport. Now it
1617 just assumes that the setting of that option is false. This bug was
1618 introduced by 4.51/PH/31.
1620 PH/28 Changed -d+all to exclude +memory, because that information is very
1621 rarely of interest, but it makes the output a lot bigger. People tend to
1622 do -d+all out of habit.
1624 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1625 code in os-type was giving problems when libc.so lives in lib64, like on
1628 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1629 aren't the modern standard, and it seems that some systems' include files
1630 don't always have them. Exim was already checking for some of the newer
1631 ones like T_AAAA, and defining it itself. I've added checks for all the
1632 record types that Exim uses.
1634 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1635 not automatically generating a new one, as it is supposed to. This
1636 prevented TLS from working. If the file did exist, but contained invalid
1637 data, a new version was generated, as expected. It was only the case of a
1638 non-existent file that was broken.
1640 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1641 with a change in libdomainkeys > 0.64.
1643 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1644 from DNS. If the selector record carries the flag, it now has
1645 precedence over the domain-wide flag.
1647 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1649 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1650 the use of a port name that isn't defined in /etc/services) occurred, the
1651 message was deferred as in a normal delivery, and thus remained on the
1652 spool, instead of being failed because of the mua_wrapper setting. This
1653 is now fixed, and I tidied up some of the mua_wrapper messages at the
1656 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1657 the messages in a hash of arrays rather than using individual hashes.
1658 This is a bit cleaner and results in dramatic memory savings, albeit
1659 at a slight CPU cost.
1661 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1662 as requested by Marc Sherman.
1664 SC/10 Eximstats: added histograms for user specified patterns as requested
1667 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1669 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1670 fopen() in the content-scanning modules that did not already have it.
1676 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1677 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1679 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1681 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1683 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1684 to test Sieve filters that use "vacation".
1686 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1687 that changes the way the GnuTLS parameters are stored in the cache file.
1688 The new format can be generated externally. For backward compatibility,
1689 if the data in the cache doesn't make sense, Exim assumes it has read an
1690 old-format file, and it generates new data and writes a new file. This
1691 means that you can't go back to an older release without removing the
1694 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1695 work if there are any delivery delays because "one_time" forces the
1696 parent to be marked "delivered", so its unseen clone is never tried
1697 again. For this reason, Exim now forbids the simultaneous setting of
1700 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1701 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1702 entry if you want to know the details. The fix, however, neglected to
1703 consider the case when local delivery batching is involved. The test for
1704 "previously delivered" was not happening when checking to see if an
1705 address could be batched with a previous (undelivered) one; under
1706 certain circumstances this could lead to multiple deliveries to the same
1709 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1710 in its include files, and this causes problems building Exim.
1712 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1713 header_syntax) but Exim was just ignoring anything given after a slash.
1714 In particular, this caused confusion with an attempt to use "verify =
1715 reverse_host_lookup/defer_ok". An error is now given when options are
1716 supplied for verify items that do not have them. (Maybe reverse_host_
1717 lookup should have a defer_ok option, but that's a different point.)
1719 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1720 defined by RFC 821) to 2048, because there were problems with some AUTH
1721 commands, and RFC 1869 says the size should be increased for extended
1722 SMTP commands that take arguments.
1724 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1727 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1728 "unknown" error; now it says that the functionality isn't in the binary.
1730 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1731 an address' error message when a string expansion fails (syntax or
1732 whatever). Otherwise the password may appear in the log. Following change
1733 PH/42 below, there is no longer a chance of it appearing in a bounce
1736 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1738 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1739 of its addresses was ever tried. (Bugzilla bug #2.)
1741 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1742 the result incorrectly in the debug output. (It correctly added a newline
1743 to what was transported.)
1745 TF/01 Added $received_time.
1747 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1748 commented out examples of how to interface to a virus scanner and to
1749 SpamAssassin. Also added commented examples of av_scanner and
1750 spamd_address settings.
1752 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1753 and controls are allowed in which ACLs. There were a couple of minor
1754 errors. Some of the entries in the conditions table (which is a table of
1755 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1756 negation of where the condition IS allowed.
1758 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1760 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1761 header file does not have a version number, so I've had to invent a new
1762 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1763 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1764 radiusclient), but it was contributed by a Radius user.
1766 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1767 files or directories, for OpenSSL.
1769 PH/19 When an Exim process that is running as root has to create an Exim log
1770 file, it does so in a subprocess that runs as exim:exim so as to get the
1771 ownership right at creation (otherwise, other Exim processes might see
1772 the file with the wrong ownership). There was no test for failure of this
1773 fork() call, which would lead to the process getting stuck as it waited
1774 for a non-existent subprocess. Forks do occasionally fail when resources
1775 run out. I reviewed all the other calls to fork(); they all seem to check
1778 PH/20 When checking for unexpected SMTP input at connect time (before writing
1779 the banner), Exim was not dealing correctly with a non-positive return
1780 from the read() function. If the client had disconnected by this time,
1781 the result was a log entry for a synchronization error with an empty
1782 string after "input=" when read() returned zero. If read() returned -1
1783 (an event I could not check), uninitialized data bytes were printed.
1784 There were reports of junk text (parts of files, etc) appearing after
1787 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1789 PH/22 Added support for macro redefinition, and (re)definition in between
1790 driver and ACL definitions.
1792 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1793 forgetting to use the resulting value; it was using the unexpanded value.
1795 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1796 hadn't been configured. The fix is from Juergen Kreileder, who
1797 understands it better than I do:
1799 "Here's what I see happening with three configured cyrus_sasl
1800 authenticators configured (plain, login, cram-md5):
1802 On startup auth_cyrus_sasl_init() gets called for each of these.
1803 This means three calls to sasl_listmech() without a specified mech_list.
1804 => SASL tests which mechs of all available mechs actually work
1805 => three warnings about OTP not working
1806 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1808 With the patch, sasl_listmech() also gets called three times. But now
1809 SASL's mech_list option is set to the server_mech specified in the the
1810 authenticator. Or in other words, the answer from sasl_listmech()
1811 gets limited to just the mech you're testing for (which is different
1813 => the return list contains just 'plain' or 'login', 'cram-md5' or
1814 nothing depending on the value of ob->server_mech.
1816 I've just tested the patch: Authentication still works fine,
1817 unavailable mechs specified in the exim configuration are still
1818 caught, and the auth.log warnings about OTP are gone."
1820 PH/25 When debugging is enabled, the contents of the command line are added
1821 to the debugging output, even when log_selector=+arguments is not
1824 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1825 answer is "GNU", and only if the return is "GNU/something" is the answer
1828 PH/27 $acl_verify_message is now set immediately after the failure of a
1829 verification in an ACL, and so is available in subsequent modifiers. In
1830 particular, the message can be preserved by coding like this:
1832 warn !verify = sender
1833 set acl_m0 = $acl_verify_message
1835 Previously, $acl_verify_message was set only while expanding "message"
1836 and "log_message" when a very denied access.
1838 PH/28 Modified OS/os.c-Linux with
1840 -#ifndef OS_LOAD_AVERAGE
1841 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1843 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1844 nomenclature these days.)
1846 PH/29 Installed patch from the Sieve maintainer that adds the options
1847 sieve_useraddress and sieve_subaddress to the redirect router.
1849 PH/30 In these circumstances:
1850 . Two addresses routed to the same list of hosts;
1851 . First host does not offer TLS;
1852 . First host accepts first address;
1853 . First host gives temporary error to second address;
1854 . Second host offers TLS and a TLS session is established;
1855 . Second host accepts second address.
1856 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1857 and peerdn, if requested) that were in fact used only for the second
1860 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1861 attention to any local part prefix or suffix that was matched by the
1862 router that accepted the address. It now behaves in the same way as it
1863 does for delivery: the affixes are removed from the local part unless
1864 rcpt_include_affixes is set on the transport.
1866 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1867 timeout during the DATA phase of an incoming message.
1869 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1870 applied a patch sanctioned by the Sieve maintainer.
1872 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1873 the uid or gid is negative. A case of a negative gid caused this to be
1874 noticed. The fix allows for either to be negative.
1876 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1877 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1880 PH/36 The change PH/12 above was broken. Fixed it.
1882 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1883 the grounds that routing the same address twice would always produce the
1884 same answer. This might have been true once, but it is certainly no
1885 longer true now. Routing a child address may depend on the previous
1886 routing that produced that child. Some complicated redirection strategies
1887 went wrong when messages had multiple recipients, and made Exim's
1888 behaviour dependent on the order in which the addresses were given.
1890 I have moved the duplicate checking until after the routing is complete.
1891 Exim scans the addresses that are assigned to local and remote
1892 transports, and removes any duplicates. This means that more work will be
1893 done, as duplicates will always all be routed, but duplicates are
1894 presumably rare, so I don't expect this is of any significance.
1896 For deliveries to pipes, files, and autoreplies, the duplicate checking
1897 still happens during the routing process, since they are not going to be
1900 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1901 It corrects a timeout issue with spamd. This is Ian's comment: "The
1902 background is that sometimes spamd either never reads data from a
1903 connection it has accepted, or it never writes response data. The exiscan
1904 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1905 blindly assumes that writes won't block so it may never time out."
1907 PH/39 Allow G after quota size as well as K and M.
1909 PH/40 The value set for $authenticated_id in an authenticator may not contain
1910 binary zeroes or newlines because the value is written to log lines and
1911 to spool files. There was no check on this. Now the value is run through
1912 the string_printing() function so that such characters are converted to
1913 printable escape sequences.
1915 PH/41 $message_linecount is a new variable that contains the total number of
1916 lines in the message. Compare $body_linecount, which is the count for the
1919 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1920 bounce and delay warning messages, except in certain special cases, which
1923 (a) An SMTP error message from a remote host;
1924 (b) A message specified in a :fail: redirection;
1925 (c) A message specified in a "fail" command in a system filter;
1926 (d) A message specified in a FAIL return from the queryprogram router;
1927 (e) A message specified by the cannot_route_message router option.
1929 In these cases only, Exim does include the error details in bounce and
1930 warning messages. There are also a few cases where bland messages such
1931 as "unrouteable address" or "local delivery error" are given.
1933 PH/43 $value is now also set for the "else" part of a ${run expansion.
1935 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1936 being worked on, but at least Exim now implements the latest version to
1939 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1940 process to complete was treated as a delivery failure, a timeout while
1941 writing the message to the pipe was logged, but erroneously treated as a
1942 successful delivery. Such timeouts include transport filter timeouts. For
1943 consistency with the overall process timeout, these timeouts are now
1944 treated as errors, giving rise to delivery failures by default. However,
1945 there is now a new Boolean option for the pipe transport called
1946 timeout_defer, which, if set TRUE, converts the failures into defers for
1947 both kinds of timeout. A transport filter timeout is now identified in
1950 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1951 systems where "make" and "gmake" are different, calling "gmake" at top
1952 level broke things. I've arranged for the value of $(MAKE) to be passed
1953 from the Makefile to this script so that it can call the same version of
1957 A note about Exim versions 4.44 and 4.50
1958 ----------------------------------------
1960 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1961 changes of various kinds. As a consequence, a big documentation update was
1962 needed. This delayed the release for rather longer than seemed good, especially
1963 in the light of a couple of (minor) security issues. Therefore, the changes
1964 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1965 release. So 4.44 and 4.50 are in effect two different branches that both start
1968 I have left the 4.50 change log unchanged; it contains all the changes since
1969 4.43. The change log for 4.44 is below; many of its items are identical to
1970 those for 4.50. This seems to be the most sensible way to preserve the
1971 historical information.
1977 1. Minor wording change to the doc/README.SIEVE file.
1979 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1980 computation of the current number of files was incorrect.
1982 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1983 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1986 4. Give more explanation in the error message when the command for a transport
1987 filter fails to execute.
1989 5. There are several places where Exim runs a non-Exim command in a
1990 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1991 was being done only for the command run by the queryprogram router. It is
1992 now done for all such subprocesses. The other cases are: ${run, transport
1993 filters, and the commands run by the lmtp and pipe transports.
1995 6. Added CONFIGURE_GROUP build-time option.
1997 7. Some older OS have a limit of 256 on the maximum number of file
1998 descriptors. Exim was using setrlimit() to set 1000 as a large value
1999 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2000 systems. I've change it so that if it can't get 1000, it tries for 256.
2002 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2003 was an oversight, and furthermore, ever since the addition of extra
2004 controls (e.g. 4.43/32), the checks on when to allow different forms of
2005 "control" were broken. There should now be diagnostics for all cases when a
2006 control that does not make sense is encountered.
2008 9. Added the /retain_sender option to "control=submission".
2010 10. $recipients is now available in the predata ACL (oversight).
2012 11. Tidy the search cache before the fork to do a delivery from a message
2013 received from the command line. Otherwise the child will trigger a lookup
2014 failure and thereby defer the delivery if it tries to use (for example) a
2015 cached ldap connection that the parent has called unbind on.
2017 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2018 of $address_data from the recipient verification was clobbered by the
2019 sender verification.
2021 13. The value of address_data from a sender verification is now available in
2022 $sender_address_data in subsequent conditions in the ACL statement.
2024 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2026 15. Added a new option "connect=<time>" to callout options, to set a different
2029 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2030 was its contents. (It was OK if the option was not defined at all.)
2032 17. A "Completed" log line is now written for messages that are removed from
2033 the spool by the -Mrm option.
2035 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2036 information about exactly what failed.
2038 19. Added -dd to debug only the daemon process.
2040 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2041 handles timeouts, both on the server side and network timeouts. Renamed the
2042 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2044 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2045 It is now set to "smtps".
2047 22. $host_address is now set to the target address during the checking of
2048 ignore_target_hosts.
2050 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2051 being passed; this would have caused $sender_host_name to have been used if
2052 matching the list had actually called for a host name (not very likely,
2053 since this list is usually IP addresses). A host name is now passed as
2056 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2057 code that specifies a non-restarting handler (typically sigaction() in
2058 modern systems) in an attempt to fix a rare and obscure crash bug.
2060 25. Narrowed the window for a race in the daemon that could cause it to ignore
2061 SIGCHLD signals. This is not a major problem, because they are used only to
2062 wake it up if nothing else does.
2064 26. A malformed maildirsize file could cause Exim to calculate negative values
2065 for the mailbox size or file count. Odd effects could occur as a result.
2066 The maildirsize information is now recalculated if the size or filecount
2069 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2070 support for a long time. Removed HAVE_SYS_VFS_H.
2072 28. Installed the latest version of exipick from John Jetmore.
2074 29. In an address list, if the pattern was not a regular expression, an empty
2075 subject address (from a bounce message) matched only if the pattern was an
2076 empty string. Non-empty patterns were not even tested. This was the wrong
2077 because it is perfectly reasonable to use an empty address as part of a
2078 database query. An empty address is now tested by patterns that are
2079 lookups. However, all the other forms of pattern expect the subject to
2080 contain a local part and a domain, and therefore, for them, an empty
2081 address still always fails if the pattern is not itself empty.
2083 30. Exim went into a mad DNS loop when attempting to do a callout where the
2084 host was specified on an smtp transport, and looking it up yielded more
2085 than one IP address.
2087 31. Re-factored the code for checking spool and log partition space into a
2088 function that finds that data and another that does the check. The former
2089 is then used to implement four new variables: $spool_space, $log_space,
2090 $spool_inodes, and $log_inodes.
2092 32. The RFC2047 encoding function was originally intended for short strings
2093 such as real names; it was not keeping to the 75-character limit for
2094 encoded words that the RFC imposes. It now respects the limit, and
2095 generates multiple encoded words if necessary. To be on the safe side, I
2096 have increased the buffer size for the ${rfc2047: expansion operator from
2099 33. It is now permitted to omit both strings after an "if" condition; if the
2100 condition is true, the result is "true". As before, when the second string
2101 is omitted, a false condition yields an empty string. This makes it less
2102 cumbersome to write custom ACL and router conditions.
2104 34. Failure to deliver a bounce message always caused it to be frozen, even if
2105 there was an errors_to setting on the router. The errors_to setting is now
2108 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2109 canonical form (fully expanded) before being placed in
2110 $sender_host_address.
2112 36. The table in the code that translates DNS record types into text (T_A to
2113 "A" for instance) was missing entries for NS and CNAME. It is just possible
2114 that this could have caused confusion if both these types were looked up
2115 for the same domain, because the text type is used as part of Exim's
2116 per-process caching. But the chance of anyone hitting this buglet seems
2119 37. The dnsdb lookup has been extended in a number of ways.
2121 (1) There is a new type, "zns", which walks up the domain tree until it
2122 finds some nameserver records. It should be used with care.
2124 (2) There is a new type, "mxh", which is like "mx" except that it returns
2125 just the host names, not the priorities.
2127 (3) It is now possible to give a list of domains (or IP addresses) to be
2128 looked up. The behaviour when one of the lookups defers can be
2129 controlled by a keyword.
2131 (4) It is now possible to specify the separator character for use when
2132 multiple records are returned.
2134 38. The dnslists ACL condition has been extended: it is now possible to supply
2135 a list of IP addresses and/or domains to be looked up in a particular DNS
2138 39. Added log_selector=+queue_time_overall.
2140 40. When running the queue in the test harness, wait just a tad after forking a
2141 delivery process, to get repeatability of debugging output.
2143 41. Include certificate and key file names in error message when GnuTLS fails
2144 to set them up, because the GnuTLS error message doesn't include the name
2145 of the failing file when there is a problem reading it.
2147 42. Allow both -bf and -bF in the same test run.
2149 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2151 44. The "Exiscan patch" is now merged into the mainline Exim source.
2153 45. Sometimes the final signoff response after QUIT could fail to get
2154 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2155 < 0 before doing a fflush(). This bug looks as though it goes back to the
2156 introduction of TLS in release 3.20, but "sometimes" must have been rare
2157 because the tests only now provoked it.
2159 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2160 (this can affect the format of dates).
2162 47. exim_tidydb, when checking for the continued existence of a message for
2163 which it has found a message-specific retry record, was not finding
2164 messages that were in split spool directories. Consequently, it was
2165 deleting retry records that should have stayed in existence.
2167 48. Steve fixed some bugs in eximstats.
2169 49. The SPA authentication driver was not abandoning authentication and moving
2170 on to the next authenticator when an expansion was forced to fail,
2171 contradicting the general specification for all authenticators. Instead it
2172 was generating a temporary error. It now behaves as specified.
2174 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2175 (the order specifies the preference for clients). The order is now AES256,
2176 AES128, 3DES, ARCFOUR128.
2178 51. Small patch to Sieve code - explicitly set From: when generating an
2181 52. Exim crashed if a remote delivery caused a very long error message to be
2182 recorded - for instance if somebody sent an entire SpamAssassin report back
2183 as a large number of 550 error lines. This bug was coincidentally fixed by
2184 increasing the size of one of Exim's internal buffers (big_buffer) that
2185 happened as part of the Exiscan merge. However, to be on the safe side, I
2186 have made the code more robust (and fixed the comments that describe what
2189 53. Now that there can be additional text after "Completed" in log lines (if
2190 the queue_time_overall log selector is set), a one-byte patch to exigrep
2191 was needed to allow it to recognize "Completed" as not the last thing in
2194 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2195 patch that reportedly fixes this has been added. I am not expert enough to
2196 create a test for it. This is what the patch creator wrote:
2198 "I found a little strange behaviour of ldap code when working with
2199 Windows 2003 AD Domain, where users was placed in more than one
2200 Organization Units. When I tried to give exim partial DN, the exit code
2201 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2202 But simultaneously result of request was absolutely normal ldap result,
2203 so I produce this patch..."
2205 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2206 REFERENCE, so I have modified the code to exclude the patch when that macro
2209 55. Some experimental protocols are using DNS PTR records for new purposes. The
2210 keys for these records are domain names, not reversed IP addresses. The
2211 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2212 leaves it alone. Component reversal etc. now happens only for IP addresses.
2215 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2217 57. Double the size of the debug message buffer (to 2048) so that more of very
2218 long debug lines gets shown.
2220 58. The exicyclog utility now does better if the number of log files to keep
2221 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2223 59. Two changes related to the smtp_active_hostname option:
2225 (1) $smtp_active_hostname is now available as a variable.
2226 (2) The default for smtp_banner uses $smtp_active_hostname instead
2227 of $primary_hostname.
2229 60. The host_aton() function is supposed to be passed a string that is known
2230 to be a valid IP address. However, in the case of IPv6 addresses, it was
2231 not checking this. This is a hostage to fortune. Exim now panics and dies
2232 if the condition is not met. A case was found where this could be provoked
2233 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2234 components; fortuitously, this particular loophole had already been fixed
2235 by change 4.50/55 above.
2237 If there are any other similar loopholes, the new check in host_aton()
2238 itself should stop them being exploited. The report I received stated that
2239 data on the command line could provoke the exploit when Exim was running as
2240 exim, but did not say which command line option was involved. All I could
2241 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2242 running as the user.
2245 61. There was a buffer overflow vulnerability in the SPA authentication code
2246 (which came originally from the Samba project). I have added a test to the
2247 spa_base64_to_bits() function which I hope fixes it.
2250 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2251 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2253 63. The daemon start-up calls getloadavg() while still root for those OS that
2254 need the first call to be done as root, but it missed one case: when
2255 deliver_queue_load_max is set with deliver_drop_privilege. This is
2256 necessary for the benefit of the queue runner, because there is no re-exec
2257 when deliver_drop_privilege is set.
2259 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2260 This has been fixed.
2262 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2263 was in use, was not putting the data itself into the right store pool;
2264 consequently, it could be overwritten for a subsequent message in the same
2265 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2268 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2270 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2271 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2272 the function were treating the return as a boolean value, which happened to
2273 work because 0=false and not-0=true, but is not correct code.
2275 68. The host_aton() function was not handling scoped IPv6 addresses (those
2276 with, for example, "%eth0" on the end) correctly.
2278 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2279 negated items (that is, ~something) in unsigned ints. Some compilers
2280 apparently mutter when there is no cast.
2282 70. If an address verification called from an ACL failed, and did not produce a
2283 user-specific message (i.e. there was only a "system" message), nothing was
2284 put in $acl_verify_message. In this situation, it now puts the system
2287 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2288 session; change 4.31/43 added the unwanted input to the log line - except
2289 that it did not do this in the start of session case. It now does.
2291 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2292 This is wrong and can cause the other end to generate a synchronization
2293 error if it is another Exim or anything else that does the synchronization
2294 check. A QUIT command is no longer sent after a timeout.
2296 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2297 during host lookups.
2299 74. The defer_ok option of callout verification was not working if it was used
2300 when verifying addresses in header lines, that is, for this case:
2302 verify = header_sender/callout=defer_ok
2304 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2305 those file descriptors could be used for SMTP connections. If anything
2306 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2307 could be sent to the SMTP client, causing chaos. The daemon now opens
2308 stdin, stdout, and stderr to /dev/null when it puts itself into the
2311 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2312 log by default. The user can override this with suitable Perl magic.
2314 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2315 the log message when discard triggers, was not working for the DATA ACL or
2316 for the non-SMTP ACL.
2318 78. Error message wording change in sieve.c.
2320 79. If smtp_accept_max_per_host was set, the number of connections could be
2321 restricted to fewer than expected, because the daemon was trying to set up
2322 a new connection before checking whether the processes handling previous
2323 connections had finished. The check for completed processes is now done
2324 earlier. On busy systems, this bug wouldn't be noticed because something
2325 else would have woken the daemon, and it would have reaped the completed
2328 80. If a message was submitted locally by a user whose login name contained one
2329 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2330 It caused a spool format error. I have fixed the spool reading code. A
2331 related problem was that the "from" clause in the Received: line became
2332 illegal because of the space(s). It is now covered by ${quote_local_part.
2334 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2337 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2339 83. Patch from Sieve maintainer to fix unterminated string problem in
2340 "vacation" handling.
2342 84. Some minor changes to the Linux configuration files to help with other
2343 OS variants using glibc.
2345 85. One more patch for Sieve to update vacation handling to latest spec.
2348 ----------------------------------------------------
2349 See the note above about the 4.44 and 4.50 releases.
2350 ----------------------------------------------------
2356 1. Change 4.43/35 introduced a bug that caused file counts to be
2357 incorrectly computed when quota_filecount was set in an appendfile
2360 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2361 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2364 3. Give more explanation in the error message when the command for a transport
2365 filter fails to execute.
2367 4. There are several places where Exim runs a non-Exim command in a
2368 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2369 was being done only for the command run by the queryprogram router. It is
2370 now done for all such subprocesses. The other cases are: ${run, transport
2371 filters, and the commands run by the lmtp and pipe transports.
2373 5. Some older OS have a limit of 256 on the maximum number of file
2374 descriptors. Exim was using setrlimit() to set 1000 as a large value
2375 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2376 systems. I've change it so that if it can't get 1000, it tries for 256.
2378 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2379 was an oversight, and furthermore, ever since the addition of extra
2380 controls (e.g. 4.43/32), the checks on when to allow different forms of
2381 "control" were broken. There should now be diagnostics for all cases when a
2382 control that does not make sense is encountered.
2384 7. $recipients is now available in the predata ACL (oversight).
2386 8. Tidy the search cache before the fork to do a delivery from a message
2387 received from the command line. Otherwise the child will trigger a lookup
2388 failure and thereby defer the delivery if it tries to use (for example) a
2389 cached ldap connection that the parent has called unbind on.
2391 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2392 of $address_data from the recipient verification was clobbered by the
2393 sender verification.
2395 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2396 was its contents. (It was OK if the option was not defined at all.)
2398 11. A "Completed" log line is now written for messages that are removed from
2399 the spool by the -Mrm option.
2401 12. $host_address is now set to the target address during the checking of
2402 ignore_target_hosts.
2404 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2405 being passed; this would have caused $sender_host_name to have been used if
2406 matching the list had actually called for a host name (not very likely,
2407 since this list is usually IP addresses). A host name is now passed as
2410 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2411 code that specifies a non-restarting handler (typically sigaction() in
2412 modern systems) in an attempt to fix a rare and obscure crash bug.
2414 15. Narrowed the window for a race in the daemon that could cause it to ignore
2415 SIGCHLD signals. This is not a major problem, because they are used only to
2416 wake it up if nothing else does.
2418 16. A malformed maildirsize file could cause Exim to calculate negative values
2419 for the mailbox size or file count. Odd effects could occur as a result.
2420 The maildirsize information is now recalculated if the size or filecount
2423 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2424 support for a long time. Removed HAVE_SYS_VFS_H.
2426 18. Updated exipick to current release from John Jetmore.
2428 19. Allow an empty sender to be matched against a lookup in an address list.
2429 Previously the only cases considered were a regular expression, or an
2432 20. Exim went into a mad DNS lookup loop when doing a callout where the
2433 host was specified on the transport, if the DNS lookup yielded more than
2436 21. The RFC2047 encoding function was originally intended for short strings
2437 such as real names; it was not keeping to the 75-character limit for
2438 encoded words that the RFC imposes. It now respects the limit, and
2439 generates multiple encoded words if necessary. To be on the safe side, I
2440 have increased the buffer size for the ${rfc2047: expansion operator from
2443 22. Failure to deliver a bounce message always caused it to be frozen, even if
2444 there was an errors_to setting on the router. The errors_to setting is now
2447 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2448 canonical form (fully expanded) before being placed in
2449 $sender_host_address.
2451 24. Updated eximstats to version 1.33
2453 25. Include certificate and key file names in error message when GnuTLS fails
2454 to set them up, because the GnuTLS error message doesn't include the name
2455 of the failing file when there is a problem reading it.
2457 26. Expand error message when OpenSSL has problems setting up cert/key files.
2460 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2461 (this can affect the format of dates).
2463 28. exim_tidydb, when checking for the continued existence of a message for
2464 which it has found a message-specific retry record, was not finding
2465 messages that were in split spool directories. Consequently, it was
2466 deleting retry records that should have stayed in existence.
2468 29. eximstats updated to version 1.35
2469 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2470 1.35 - bugfix such that pie charts by volume are generated correctly
2472 30. The SPA authentication driver was not abandoning authentication and moving
2473 on to the next authenticator when an expansion was forced to fail,
2474 contradicting the general specification for all authenticators. Instead it
2475 was generating a temporary error. It now behaves as specified.
2477 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2478 (the order specifies the preference for clients). The order is now AES256,
2479 AES128, 3DES, ARCFOUR128.
2481 31. Small patch to Sieve code - explicitly set From: when generating an
2484 32. Exim crashed if a remote delivery caused a very long error message to be
2485 recorded - for instance if somebody sent an entire SpamAssassin report back
2486 as a large number of 550 error lines. This bug was coincidentally fixed by
2487 increasing the size of one of Exim's internal buffers (big_buffer) that
2488 happened as part of the Exiscan merge. However, to be on the safe side, I
2489 have made the code more robust (and fixed the comments that describe what
2492 33. Some experimental protocols are using DNS PTR records for new purposes. The
2493 keys for these records are domain names, not reversed IP addresses. The
2494 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2495 leaves it alone. Component reversal etc. now happens only for IP addresses.
2498 34. The host_aton() function is supposed to be passed a string that is known
2499 to be a valid IP address. However, in the case of IPv6 addresses, it was
2500 not checking this. This is a hostage to fortune. Exim now panics and dies
2501 if the condition is not met. A case was found where this could be provoked
2502 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2503 components; fortuitously, this particular loophole had already been fixed
2504 by change 4.50/55 or 4.44/33 above.
2506 If there are any other similar loopholes, the new check in host_aton()
2507 itself should stop them being exploited. The report I received stated that
2508 data on the command line could provoke the exploit when Exim was running as
2509 exim, but did not say which command line option was involved. All I could
2510 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2511 running as the user.
2514 35. There was a buffer overflow vulnerability in the SPA authentication code
2515 (which came originally from the Samba project). I have added a test to the
2516 spa_base64_to_bits() function which I hope fixes it.
2519 36. The daemon start-up calls getloadavg() while still root for those OS that
2520 need the first call to be done as root, but it missed one case: when
2521 deliver_queue_load_max is set with deliver_drop_privilege. This is
2522 necessary for the benefit of the queue runner, because there is no re-exec
2523 when deliver_drop_privilege is set.
2525 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2526 was in use, was not putting the data itself into the right store pool;
2527 consequently, it could be overwritten for a subsequent message in the same
2528 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2531 38. Sometimes the final signoff response after QUIT could fail to get
2532 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2533 < 0 before doing a fflush(). This bug looks as though it goes back to the
2534 introduction of TLS in release 3.20, but "sometimes" must have been rare
2535 because the tests only now provoked it.
2541 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2542 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2543 Now it returns an integer. A number of calls were still expecting a T/F
2544 return. Fortuitously, in all cases, the tests worked in OK situations,
2545 which is the norm. However, things would have gone wrong on any write
2546 failures on the smtp file descriptor. This function is used when sending
2547 messages over SMTP and also when doing verify callouts.
2549 2. When Exim is called to do synchronous delivery of a locally submitted
2550 message (the -odf or -odi options), it no longer closes stderr before doing
2553 3. Implemented the mua_wrapper option.
2555 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2557 5. Implemented the functions header_remove(), header_testname(),
2558 header_add_at_position(), and receive_remove_recipient(), and exported them
2561 6. If an ACL "warn" statement specified the addition of headers, Exim already
2562 inserted X-ACL-Warn: at the start if there was no header name. However, it
2563 was not making this test for the second and subsequent header lines if
2564 there were newlines in the string. This meant that an invalid header could
2565 be inserted if Exim was badly configured.
2567 7. Allow an ACL "warn" statement to add header lines at the start or after all
2568 the Received: headers, as well as at the end.
2570 8. Added the rcpt_4xx retry error code.
2572 9. Added postmaster_mailfrom=xxx to callout verification option.
2574 10. Added mailfrom=xxxx to the callout verification option, for verify=
2577 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2578 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2580 12. Inserted some casts to stop certain compilers warning when using pointer
2581 differences as field lengths or precisions in printf-type calls (mostly
2582 affecting debugging statements).
2584 13. Added optional readline() support for -be (dynamically loaded).
2586 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2587 same clock tick as a message's arrival, so that its received time was the
2588 same as the "first fail" time on the retry record, and that message
2589 remained on the queue past the ultimate address timeout, every queue runner
2590 would try a delivery (because it was past the ultimate address timeout) but
2591 after another failure, the ultimate address timeout, which should have then
2592 bounced the address, did not kick in. This was a "< instead of <=" error;
2593 in most cases the first failure would have been in the next clock tick
2594 after the received time, and all would be well.
2596 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2597 being recognized when the domain list was tested by the match_domain
2598 condition in an expansion string.
2600 16. Added the ${str2b64: operator.
2602 17. Exim was always calling setrlimit() to set a large limit for the number of
2603 processes, without checking whether the existing limit was already
2604 adequate. (It did check for the limit on file descriptors.) Furthermore,
2605 errors from getrlimit() and setrlimit() were being ignored. Now they are
2606 logged to the main and panic logs, but Exim does carry on, to try to do its
2607 job under whatever limits there are.
2609 18. Imported PCRE 5.0.
2611 19. Trivial typo in log message " temporarily refused connection" (the leading
2614 20. If the log selector return_path_on_delivery was set and an address was
2615 redirected to /dev/null, the delivery process crashed because it assumed
2616 that a return path would always be set for a "successful" delivery. In this
2617 case, the whole delivery is bypassed as an optimization, and therefore no
2620 21. Internal re-arrangement: the function for sending a challenge and reading
2621 a response while authentication was assuming a zero-terminated challenge
2622 string. It's now changed to take a pointer and a length, to allow for
2623 binary data in such strings.
2625 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2627 23. Exim was not respecting finduser_retries when seeking the login of the
2628 uid under which it was called; it was always trying 10 times. (The default
2629 setting of finduser_retries is zero.) Also, it was sleeping after the final
2630 failure, which is pointless.
2632 24. Implemented tls_on_connect_ports.
2634 25. Implemented acl_smtp_predata.
2636 26. If the domain in control=submission is set empty, Exim assumes that the
2637 authenticated id is a complete email address when it generates From: or
2638 Sender: header lines.
2640 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2641 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2642 chown and chgrp in /bin and hostname in /usr/bin.
2644 28. Exim was keeping the "process log" file open after each use, just as it
2645 does for the main log. This opens the possibility of it remaining open for
2646 long periods when the USR1 signal hits a daemon. Occasional processlog
2647 errors were reported, that could have been caused by this. Anyway, it seems
2648 much more sensible not to leave this file open at all, so that is what now
2651 29. The long-running daemon process does not normally write to the log once it
2652 has entered its main loop, and it closes the log before doing so. This is
2653 so that log files can straightforwardly be renamed and moved. However,
2654 there are a couple of unusual error situations where the daemon does write
2655 log entries, and I had neglected to close the log afterwards.
2657 30. The text of an SMTP error response that was received during a remote
2658 delivery was being truncated at 512 bytes. This is too short for some of
2659 the long messages that one sometimes sees. I've increased the limit to
2662 31. It is now possible to make retry rules that apply only when a message has a
2663 specific sender, in particular, an empty sender.
2665 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2666 it possible to be selective about when SMTP synchronization is enforced.
2668 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2670 32. Implemented hosts_connection_nolog.
2672 33. Added an ACL for QUIT.
2674 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2677 35. Added mailbox_size and mailbox_filecount to appendfile.
2679 36. Added control = no_multiline_responses to ACLs.
2681 37. There was a bug in the logic of the code that waits for the clock to tick
2682 in the case where the clock went backwards by a substantial amount such
2683 that the microsecond fraction of "now" was more than the microsecond
2684 fraction of "then" (but the whole seconds number was less).
2686 38. Added support for the libradius Radius client library this is found on
2687 FreeBSD (previously only the radiusclient library was supported).
2693 1. When certain lookups returned multiple values in the form name=value, the
2694 quoting of the values was not always being done properly. Specifically:
2695 (a) If the value started with a double quote, but contained no whitespace,
2697 (b) If the value contained whitespace other than a space character (i.e.
2698 tabs or newlines or carriage returns) it was not quoted.
2699 This fix has been applied to the mysql and pgsql lookups by writing a
2700 separate quoting function and calling it from the lookup code. The fix
2701 should probably also be applied to nisplus, ibase and oracle lookups, but
2702 since I cannot test any of those, I have not disturbed their existing code.
2704 2. A hit in the callout cache for a specific address caused a log line with no
2705 reason for rejecting RCPT. Now it says "Previous (cached) callout
2706 verification failure".
2708 3. There was an off-by-one bug in the queryprogram router. An over-long
2709 return line was truncated at 256 instead of 255 characters, thereby
2710 overflowing its buffer with the terminating zero. As well as fixing this, I
2711 have increased the buffer size to 1024 (and made a note to document this).
2713 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2714 when Exim is waiting for an SMTP response from a remote server, Exim
2715 restarts its select() call on the socket, thereby resetting its timeout.
2716 This is not a problem when such interrupts are rare. Somebody set up a cron
2717 job to run exiwhat every 2 minutes, which is less than the normal select()
2718 timeout (5 or 10 minutes). This meant that the select() timeout never
2719 kicked in because it was always reset. I have fixed this by comparing the
2720 time when an interrupt arrives with the time at the start of the first call
2721 to select(). If more time than the timeout has elapsed, the interrupt is
2722 treated as a timeout.
2724 5. Some internal re-factoring in preparation for the addition of Sieve
2725 extensions (by MH). In particular, the "personal" test is moved to a
2726 separate function, and given an option for scanning Cc: and Bcc: (which is
2727 not set for Exim filters).
2729 6. When Exim created an email address using the login of the caller as the
2730 local part (e.g. when creating a From: or Sender: header line), it was not
2731 quoting the local part when it contained special characters such as @.
2733 7. Installed new OpenBSD configuration files.
2735 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2736 try to make them clearer.
2738 9. Callout options, other than the timeout value, were being ignored when
2739 verifying sender addresses in header lines. For example, when using
2741 verify = header_sender/callout=no_cache
2743 the cache was (incorrectly) being used.
2745 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2746 only the Cygwin environment.
2748 11. When return_path_on_delivery was set as a log selector, if different remote
2749 addresses in the same message used different return paths and parallel
2750 remote delivery occurred, the wrong values would sometimes be logged.
2751 (Whenever a remote delivery process finished, the return path value from
2752 the most recently started remote delivery process was logged.)
2754 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2755 lines when AUTH and/or TLS are in use. This is the "received protocol"
2756 field. Exim used to use "asmtp" for authenticated SMTP, without any
2757 indication (in the protocol name) for TLS use. Now it follows the RFC and
2758 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2759 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2760 names appear in log lines as well as in Received: header lines.
2762 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2763 extensions, and comparison tests, and to fix some bugs.
2765 14. Changes to the "personal" filter test:
2767 (1) The test was buggy in that it was just doing the equivalent of
2768 "contains" tests on header lines. For example, if a user's address was
2769 anne@some.where, the "personal" test would incorrectly be true for
2771 To: susanne@some.where
2773 This test is now done by extracting each address from the header in turn,
2774 and checking the entire address. Other tests that are part of "personal"
2775 are now done using regular expressions (for example, to check local parts
2776 of addresses in From: header lines).
2778 (2) The list of non-personal local parts in From: addresses has been
2779 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2780 taken from the Sieve specification recommendations.
2782 (3) If the message contains any header line starting with "List-" it is
2783 treated as non-personal.
2785 (4) The test for "circular" in the Subject: header line has been removed
2786 because it now seems ill-conceived.
2788 15. Minor typos in src/EDITME comments corrected.
2790 16. Installed latest exipick from John Jetmore.
2792 17. If headers_add on a router specified a text string that was too long for
2793 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2794 of string_sprintf() is now avoided.
2796 18. $message_body_size was not set (it was always zero) when running the DATA
2797 ACL and the local_scan() function.
2799 19. For the "mail" command in an Exim filter, no default was being set for
2800 the once_repeat time, causing a random time value to be used if "once" was
2801 specified. (If the value happened to be <= 0, no repeat happened.) The
2802 default is now 0s, meaning "never repeat". The "vacation" command was OK
2803 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2804 (I found it when inspecting the code).
2806 20. There is now an overall timeout for performing a callout verification. It
2807 defaults to 4 times the callout timeout, which applies to individual SMTP
2808 commands during the callout. The overall timeout applies when there is more
2809 than one host that can be tried. The timeout is checked before trying the
2810 next host. This prevents very long delays if there are a large number of
2811 hosts and all are timing out (e.g. when the network connections are timing
2812 out). The value of the overall timeout can be changed by specifying an
2813 additional sub-option for "callout", called "maxwait". For example:
2815 verify = sender/callout=5s,maxwait=20s
2817 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2818 to the end before writing, but this should make it even safer).
2820 22. Exim was forgetting that it had advertised PIPELINING for the second and
2821 subsequent messages on an SMTP connection. It was also not resetting its
2822 memory on STARTTLS and an internal HELO.
2824 23. When Exim logs an SMTP synchronization error within a session, it now
2825 records whether PIPELINING has been advertised or not.
2827 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2828 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2829 rather than long int.
2831 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2833 26. Added the never_mail option to autoreply.
2839 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2840 crash if the getsockname() call failed; this can happen if a connection is
2841 closed very soon after it is established. The problem was simply in the
2842 order in which certain operations were done, causing Exim to try to write
2843 to the SMTP stream before it had set up the file descriptor. The bug has
2844 been fixed by making things happen in the correct order.
2850 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2851 before the connection was closed, thus losing the rejection response.
2853 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2854 some early Solaris releases, but causes trouble in current releases where
2855 socklen_t is defined.
2857 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2860 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2863 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2864 that releasing the top of it at the end releases what was used for sub-
2865 expansions (unless the block got too big). However, discard this block if
2866 the first thing is a variable or header, so that we can use its block when
2867 it is dynamic (useful for very large $message_headers, for example).
2869 6. Lookups now cache *every* query, not just the most recent. A new, separate
2870 store pool is used for this. It can be recovered when all lookup caches are
2871 flushed. Lookups now release memory at the end of their result strings.
2872 This has involved some general refactoring of the lookup sources.
2874 7. Some code has been added to the store_xxx() functions to reduce the amount
2875 of flapping under certain conditions.
2877 8. log_incoming_interface used to affect only the <= reception log lines. Now
2878 it causes the local interface and port to be added to several more SMTP log
2879 lines, for example "SMTP connection from", and rejection lines.
2881 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2883 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2885 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2886 could be overwritten at the end of the current message (or the start of a
2887 new message if it was set in a HELO ACL). The value is now preserved for
2888 the duration of the SMTP connection.
2890 12. If a transport had a headers_rewrite setting, and a matching header line
2891 contained an unqualified address, that address was qualified, even if it
2892 did not match any rewriting rules. The underlying bug was that the values
2893 of the flags that permit the existence of unqualified sender and recipient
2894 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2895 non-local messages, and by -bnq for local messages) were not being
2896 preserved with the message after it was received.
2898 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2899 "next input=" as part of the text comprising the host identity instead of
2900 the correct text. The code was using the same buffer for two different
2901 strings. However, depending on which order the printing function evaluated
2902 its arguments, the bug did not always show up. Under Linux, for example, my
2903 test suite worked just fine.
2905 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2906 doesn't work with some older versions of Perl. It has been changed to "my",
2907 which in any case is probably the better facility to use.
2909 15. A really picky compiler found some instances of statements for creating
2910 error messages that either had too many or two few arguments for the format
2913 16. The size of the buffer for calls to the DNS resolver has been increased
2914 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2915 for addresses that have a lot of PTR records. This alleviates a problem; it
2916 does not fully solve it.
2918 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2919 buffer now truncates the list and logs the incident, which is the same
2920 action as happens when Exim is looking up a host name and its aliases.
2921 Previously in this situation something unpredictable would happen;
2922 sometimes it was "internal error: store_reset failed".
2924 18. If a server dropped the connection unexpectedly when an Exim client was
2925 using GnuTLS and trying to read a response, the client delivery process
2926 crashed while trying to generate an error log message.
2928 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2929 string, for example:
2931 warn message = H1: something\nH2: something
2933 the text was added as a single header line from Exim's point of view
2934 though it ended up OK in the delivered message. However, searching for the
2935 second and subsequent header lines using $h_h2: did not work. This has been
2936 fixed. Similarly, if a system filter added multiple headers in this way,
2937 the routers could not see them.
2939 20. Expanded the error message when iplsearch is called with an invalid key to
2940 suggest using net-iplsearch in a host list.
2942 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2943 ACLs are no longer actually imposed (and a message to that effect is
2946 22. If a "gecos" field in a passwd entry contained escaped characters, in
2947 particular, if it contained a \" sequence, Exim got it wrong when building
2948 a From: or a Sender: header from that name. A second bug also caused
2949 incorrect handling when an unquoted " was present following a character
2950 that needed quoting.
2952 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2953 was not being matched caselessly.
2955 24. Arranged for all hyphens in the exim.8 source to be escaped with
2958 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2959 properly. Recipient callout cache records were still being keyed to include
2960 the sender, even when use_sender was set false. This led to far more
2961 callouts that were necessary. The sender is no longer included in the key
2962 when use_sender is false.
2964 26. Added "control = submission" modifier to ACLs.
2966 27. Added the ${base62d: operator to decode base 62 numbers.
2968 28. dnsdb lookups can now access SRV records.
2970 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2971 the configuration file.
2973 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2974 (-v) mode. This makes the output for a verbose queue run more intelligible.
2976 31. Added a use_postmaster feature to recipient callouts.
2978 32. Added the $body_zerocount variable, containing the number of binary zero
2979 bytes in the message body.
2981 33. The time of last modification of the "new" subdirectory is now used as the
2982 "mailbox time last read" when there is a quota error for a maildir
2985 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2987 35. Added +ignore_unknown as a special item in host lists.
2989 36. Code for decoding IPv6 addresses in host lists is now included, even if
2990 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2991 address was recognized as an IP address, but was then not correctly decoded
2992 into binary, causing unexpected and incorrect effects when compared with
2999 1. Very minor rewording of debugging text in manualroute to say "list of
3000 hosts" instead of "hostlist".
3002 2. If verify=header_syntax was set, and a header line with an unqualified
3003 address (no domain) and a large number of spaces between the end of the
3004 name and the colon was received, the reception process suffered a buffer
3005 overflow, and (when I tested it) crashed. This was caused by some obsolete
3006 code that should have been removed. The fix is to remove it!
3008 3. When running in the test harness, delay a bit after writing a bounce
3009 message to get a bit more predictability in the log output.
3011 4. Added a call to search_tidyup() just before forking a reception process. In
3012 theory, someone could use a lookup in the expansion of smtp_accept_max_
3013 per_host which, without the tidyup, could leave open a database connection.
3015 5. Added the variables $recipient_data and $sender_data which get set from a
3016 lookup success in an ACL "recipients" or "senders" condition, or a router
3017 "senders" option, similar to $domain_data and $local_part_data.
3019 6. Moved the writing of debug_print from before to after the "senders" test
3022 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3023 problems for message scanning, either using a data ACL, or using
3024 local_scan() because the Received: header was not generated till after they
3025 were called (in order to set the time as the time of reception completion).
3026 I have revised the way this works. The header is now generated after the
3027 body is received, but before the ACL or local_scan() are called. After they
3028 are run, the timestamp in the header is updated.
3034 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3035 before starting a queue runner without re-exec. This happened only when
3036 deliver_drop_privilege was set or when the Exim user was set to root. The
3037 effect of the bug was that timeouts during subsequent deliveries caused
3038 crashes instead of being properly handled. The handler is now left at its
3039 default (and expected) setting.
3041 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3042 message, again when deliver_drop_privilege is set or Exim is run as root.
3043 The bug described in (1) was not present in this case, but the tidying up
3044 of the other signals was missing. I have made the two cases consistent.
3046 3. The ignore_target_hosts setting on a manualroute router was being ignored
3047 for hosts that were looked up using the /MX notation.
3049 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3052 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3053 operated on the sender address. After changing the $sender_address to <>
3054 for the sender address verify, Exim was re-instated it as the original
3055 (before rewriting) address, but remembering that it had rewritten it, so it
3056 wasn't rewriting it again. This bug also had the effect of breaking the
3057 sender address verification caching when the sender address was rewritten.
3059 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3060 This has been changed so that if the ip literal address matches
3061 ignore_target_hosts, the router declines.
3063 7. Added expansion conditions match_domain, match_address, and match_local_
3064 part (NOT match_host).
3066 8. The placeholder for the Received: header didn't have a length field set.
3068 9. Added code to Exim itself and to exim_lock to test for a specific race
3069 condition that could lead to file corruption when using MBX delivery. The
3070 issue is with the lockfile that is created in /tmp. If this file is removed
3071 after a process has opened it but before that process has acquired a lock,
3072 there is the potential for a second process to recreate the file and also
3073 acquire a lock. This could lead to two Exim processes writing to the file
3074 at the same time. The added code performs the same test as UW imapd; it
3075 checks after acquiring the lock that its file descriptor still refers to
3076 the same named file.
3078 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3079 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3082 11. Added the smtp_active_hostname option. If used, this will typically be made
3083 to depend on the incoming interface address. Because $interface_address is
3084 not set up until the daemon has forked a reception process, error responses
3085 that can happen earlier (such as "too many connections") no longer contain
3088 12. If an expansion in a condition on a "warn" statement fails because a lookup
3089 defers, the "warn" statement is abandoned, and the next ACL statement is
3090 processed. Previously this caused the whole ACL to be aborted.
3092 13. Added the iplsearch lookup type.
3094 14. Added ident_timeout as a log selector.
3096 15. Added tls_certificate_verified as a log selector.
3098 16. Added a global option tls_require_ciphers (compare the smtp transport
3099 option of the same name). This controls incoming TLS connections.
3101 17. I finally figured out how to make tls_require_ciphers do a similar thing
3102 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3103 before starting the TLS session.
3105 18. Tabs are now shown as \t in -bP output.
3107 19. If the log selector return_path_on_delivery was set, Exim crashed when
3108 bouncing a message because it had too many Received: header lines.
3110 20. If two routers both had headers_remove settings, and the first one included
3111 a superfluous trailing colon, the final name in the first list and the
3112 first name in the second list were incorrectly joined into one item (with a
3113 colon in the middle).
3119 1. Added -C and -D options to the exinext utility, mainly to make it easier
3120 to include in the automated testing, but these could be helpful when
3121 multiple configurations are in use.
3123 2. The exinext utility was not formatting the output nicely when there was
3124 an alternate port involved in the retry record key, nor when there was a
3125 message id as well (for retries that were specific to a specific message
3126 and a specific host). It was also confused by IPv6 addresses, because of
3127 the additional colons they contain. I have fixed the IPv4 problem, and
3128 patched it up to do a reasonable job for IPv6.
3130 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3131 delivery, the log line now contains "pipelined" if PIPELINING was used.
3133 4. An SMTP transport process used to panic and die if the bind() call to set
3134 an explicit outgoing interface failed. This has been changed; it is now
3135 treated in the same way as a connect() failure.
3137 5. A reference to $sender_host_name in the part of a conditional expansion
3138 that was being skipped was still causing a DNS lookup. This no longer
3141 6. The def: expansion condition was not recognizing references to header lines
3142 that used bh_ and bheader_.
3144 7. Added the _cache feature to named lists.
3146 8. The code for checking quota_filecount in the appendfile transport was
3147 allowing one more file than it should have been.
3149 9. For compatibility with Sendmail, the command line option
3157 and sets the incoming protocol and host name (for trusted callers). The
3158 host name and its colon can be omitted when only the protocol is to be set.
3159 Note the Exim already has two private options, -pd and -ps, that refer to
3160 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3161 "s", but I don't think that's a major issue.
3163 10. A number of refactoring changes to the code, none of which should affect
3166 (a) The number of logging options was getting close to filling up the
3167 32-bit word that was used as a bit map. I have split them into two classes:
3168 those that are passed in the argument to log_write(), and those that are
3169 only ever tested independently outside of that function. These are now in
3170 separate 32-bit words, so there is plenty of room for expansion again.
3171 There is no change in the user interface or the logging behaviour.
3173 (b) When building, for example, log lines, the code previously used a
3174 macro that called string_cat() twice, in order to add two strings. This is
3175 not really sufficiently general. Furthermore, there was one instance where
3176 it was actually wrong because one of the argument was used twice, and in
3177 one call a function was used. (As it happened, calling the function twice
3178 did not affect the overall behaviour.) The macro has been replaced by a
3179 function that can join an arbitrary number of extra strings onto a growing
3182 (c) The code for expansion conditions now uses a table and a binary chop
3183 instead of a serial search (which was left over from when there were very
3184 few conditions). Also, it now recognizes conditions like "pam" even when
3185 the relevant support is not compiled in: a suitably worded error message is
3186 given if an attempt is made to use such a condition.
3188 11. Added ${time_interval:xxxxx}.
3190 12. A bug was causing one of the ddress fields not to be passed back correctly
3191 from remote delivery subprocesses. The field in question was not being
3192 subsequently used, so this caused to problems in practice.
3194 13. Added new log selectors queue_time and deliver_time.
3196 14. Might have fixed a bug in maildirsizefile handling that threw up
3197 "unexpected character" debug warnings, and recalculated the data
3198 unnecessarily. In any case, I expanded the warning message to give more
3201 15. Added the message "Restricted characters in address" to the statements in
3202 the default ACL that block characters like @ and % in local parts.
3204 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3205 Three changes have been made:
3207 (a) There was a serious bug; a negative response to MAIL caused the whole
3208 recipient domain to be cached as invalid, thereby blocking all messages
3209 to all local parts at the same domain, from all senders. This bug has
3210 been fixed. The domain is no longer cached after a negative response to
3211 MAIL if the sender used is not empty.
3213 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3216 (c) A new callout option, "use_sender" has been added for people who want
3217 the modified behaviour.
3223 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3226 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3227 indeed breaks things for older releases.
3229 3. Added additional logging to the case where there is a problem reading data
3230 from a filter that is running in a subprocess using a pipe, in order to
3231 try to track down a specific problem.
3233 4. Testing facility fudge: when running in the test harness and attempting
3234 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3235 getting "No route to host". Convert this to a timeout.
3237 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3240 6. Some OS don't have socklen_t but use size_t instead. This affects the
3241 fifth argument of getsockopt() amongst other things. This is now
3242 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3243 can be set for individual OS. I have set it for SunOS5, OSF1, and
3244 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3245 some earlier ones do not.
3247 7. Change 4.30/15 was not doing the test caselessly.
3249 8. The standard form for an IPv6 address literal was being rejected by address
3250 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3251 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3252 this, as well as the form without the "IPv6" on the front (but only when
3253 address literals are enabled, of course).
3255 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3257 10. Exim crashed if a message with an empty sender address specified by -f
3258 encountered a router with an errors_to setting. This could be provoked only
3259 by a command such as
3263 where an empty string was supplied; "<>" did not hit this bug.
3265 11. Installed PCRE release 4.5.
3267 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3268 remained set. It is now erased.
3270 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3271 times from message ids (which are base 36 rather than the normal 62).
3273 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3274 were being counted as actual protocol errors, and logged if the log
3275 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3276 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3277 and DATA following a set of rejected RCPTs do not count as protocol errors.
3278 In other words, Exim assumes they were pipelined, though this may not
3279 actually be the case. Of course, in all cases the client gets an
3280 appropriate error code.
3282 15. If a lookup fails in an ACL condition, a message about the failure may
3283 be available; it is used if testing the ACL cannot continue, because most
3284 such messages specify what the cause of the deferral is. However, some
3285 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3286 that caused an old message to be retained and used if a later statement
3287 caused a defer, replacing the real cause of the deferral.
3289 16. If an IP address had so many PTR records that the DNS lookup buffer
3290 was not large enough to hold them, Exim could crash while trying to process
3291 the truncated data. It now detects and logs this case.
3293 17. Further to 4.21/58, another change has been made: if (and only if) the
3294 first line of a message (the first header line) ends with CRLF, a bare LF
3295 in a subsequent header line has a space inserted after it, so as not to
3296 terminate the header.
3298 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3299 unnecessarily, used atoi() instead of strtol(), and didn't check the
3300 termination when getting file sizes from file names by regex.
3302 19. Completely re-implemented the support for maildirsize files, in the light
3303 of a number of problems with the previous contributed implementation
3304 (4.30/29). In particular:
3306 . If the quota is zero, the maildirsize file is maintained, but no quota is
3309 . If the maildir directory does not exist, it is created before any attempt
3310 to write a maildirsize file.
3312 . The quota value in the file is just a cache; if the quota is changed in
3313 the transport, the new value overrides.
3315 . A regular expression is available for excluding directories from the
3318 20. The autoreply transport checks the characters in options that define the
3319 message's headers; it allows continued headers, but it was checking with
3320 isspace() after an embedded newline instead of explicitly looking for a
3323 21. If all the "regular" hosts to which an address was routed had passed their
3324 expiry times, and had not reached their retry times, the address was
3325 bounced, even if fallback hosts were defined. Now Exim should go on to try
3328 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3329 equivalent code in the SMTP transport. Some hosts send humungous responses
3330 to HELO/EHLO, more than 1024 it seems.
3332 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3333 gives compiler warnings in some environments. I've now done it "properly",
3336 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3337 (because of problems with the built-in one) was declared to return uschar *
3338 instead of char *, causing compiler failure.
3340 25. Fixed a file descriptor leak when processing alias/forward files.
3342 26. Fixed a minor format string issue in dbfn.c.
3344 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3346 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3347 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3350 29. When a sender address is verified, it is cached, to save repeating the test
3351 when there is more than one recipient in a message. However, when the
3352 verification involves a callout, it is possible for different callout
3353 options to be set for different recipients. It is too complicated to keep
3354 track of this in the cache, so now Exim always runs a verification when a
3355 callout is required, relying on the callout cache for the optimization.
3356 The overhead is duplication of the address routing, but this should not be
3359 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3360 to be verified with callout=postmaster, and the main callout worked but the
3361 postmaster check failed, the verification correctly failed. However, if a
3362 subsequent RCPT command asked for sender verification *without* the
3363 postmaster check, incorrect caching caused this verification also to fail,
3366 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3367 it was not caching the DNS options (qualify_single, search_parents) that
3368 were used when the lookup failed. A subsequent lookup with different
3369 options therefore always gave the same answer, though there were cases
3370 where it should not have. (Example: a "domains = !$mx_any" option on a
3371 dnslookup router: the "domains" option is always processed without any
3372 widening, but the router might have qualify_single set.) Now Exim uses the
3373 cached value only when the same options are set.
3375 32. Added John Jetmore's "exipick" utility to the distribution.
3377 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3378 than a timeout (e.g. a certificate is required, and is not provided), an
3379 Exim server now closes the connection immediately. Previously it waited for
3380 the client to close - but if the client is SSL, it seems that they each
3381 wait for each other, leading to a delay before one of them times out.
3383 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3384 maintained 0.8.x compatibility because I don't think many are using it, and
3385 it is clearly obsolete.
3387 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3390 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3391 client certificate was expired. A simple patch fixes this, though I don't
3392 understand the full logic of why the verify callback is called multiple
3395 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3396 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3397 which causes problems with some clients (such as the Certicom SSL Plus
3398 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3399 disables the coutermeasure allowing Eudora to connect."
3401 38. Exim was not checking that a write() to a log file succeeded. This could
3402 lead to Bad Things if a log got too big, in particular if it hit a file
3403 size limit. Exim now panics and dies if it cannot write to a log file, just
3404 as it does if it cannot open a log file.
3406 39. Modified OS/Makefile-Linux so that it now contains
3408 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3410 The two -D definitions ensure that Exim is compiled with large file
3411 support, which makes it possible to handle log files that are bigger than
3414 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3415 instance) a domain was checked against a named list that involved a lookup,
3416 causing $domain_data to be set, then another domain was checked against the
3417 same list, then the first domain was re-checked, the value of $domain_data
3418 after the final check could be wrong. In particular, if the second check
3419 failed, it could be set empty. This bug probably also applied to
3422 41. The strip_trailing_dot option was not being applied to the address given
3423 with the -f command-line option.
3425 42. The code for reading a message's header from the spool was incrementing
3426 $received_count, but never initializing it. This meant that the value was
3427 incorrect (doubled) while delivering a message in the same process in which
3428 it was received. In the most common configuration of Exim, this never
3429 happens - a fresh exec is done - but it can happen when
3430 deliver_drop_privilege is set.
3432 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3433 it now includes up to 150 characters of the unexpected data in the log
3436 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3437 and building data strings. The size of both of these buffers was 10 000
3438 bytes - far larger than anybody would *ever* want, thought I. Needless to
3439 say, somebody hit the limit. I have increased the maximum line length to
3440 20 000 and the maximum data length of concatenated lines to 100 000. I have
3441 also fixed two bugs, because there was no checking on these buffers. Tsk,
3442 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3443 buffer is too small.
3445 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3446 lsearch lookups. Now it does.
3448 46. When parsing a route_list item in a manualroute router, a fixed-length
3449 buffer was used for the list of hosts. I made this 1024 bytes long,
3450 thinking that nobody would ever have a list of hosts that long. Wrong.
3451 Somebody had a whole pile of complicated expansion conditions, and the
3452 string was silently truncated, leading to an expansion error. It turns out
3453 that it is easier to change to an unlimited length (owing to other changes
3454 that have happened since this code was originally written) than to build
3455 structure for giving a limitation error. The length of the item that
3456 expands into the list of hosts is now unlimited.
3458 47. The lsearch lookup could not handle data where the length of text line was
3459 more than 4095 characters. Such lines were truncated, leading to shortened
3460 data being returned. It should now handle lines of any length.
3462 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3463 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3466 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3467 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3468 no longer happens in comment lines. A list of replacements is now placed
3469 at the head of all of the source files, except those whose only change is
3470 to replace PERL_COMMAND in the very first #! line.
3472 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3473 messages on the queue, with a bottom-up merge sort, using code contributed
3474 by Michael Haardt. This should make operations like -bp somewhat faster on
3475 large queues. It won't affect queue runners, except when queue_run_in_order
3478 51. Installed eximstats 1.31 in the distribution.
3480 52. Added support for SRV lookups to the dnslookup router.
3482 53. If an ACL referred to $message_body or $message_body_end, the value was not
3483 reset for any messages that followed in the same SMTP session.
3485 54. The store-handling optimization for building very long strings was not
3486 differentiating between the different store pools. I don't think this
3487 actually made any difference in practice, but I've tidied it.
3489 55. While running the routers to verify a sender address, $sender_address
3490 was still set to the sender address. This is wrong, because when routing to
3491 send a bounce to the sender, it would be empty. Therefore, I have changed
3492 it so that, while verifying a sender address, $sender_address is set to <>.
3493 (There is no change to what happens when verifying a recipient address.)
3495 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3496 target A or AAAA records (if not already returned) without resetting the
3497 qualify_single or search_parents options of the DNS resolver. These are
3498 inappropriate in this case because the targets of MX and SRV records must
3499 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3500 target that, when qualified, matched something in the local domain. These
3501 two options are now turned off when doing these lookups.
3503 57. It seems that at least some releases of Reiserfs (which does not have the
3504 concept of a fixed number of inodes) returns zero and not -1 for the
3505 number of available inodes. This interacted badly with check_spool_inodes,
3506 which assumed that -1 was the "no such thing" setting. What I have done is
3507 to check that the total number of inodes is greater than zero before doing
3508 the test of how many are available.
3510 58. When a "warn" ACL statement has a log_message modifier, the message is
3511 remembered, and not repeated. This is to avoid a lot of repetition when a
3512 message has many recipients that cause the same warning to be written.
3513 Howewer, Exim was preserving the list of already written lines for an
3514 entire SMTP session, which doesn't seem right. The memory is now reset if a
3515 new message is started.
3517 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3518 debugging output unless log_rewrite was also set.
3520 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3521 of (int)(handle) when we know that handle contains (void *)(-1).
3523 61. The Exim daemon panic-logs an error return when it closes the incoming
3524 connection. However "connection reset by peer" seems to be common, and
3525 isn't really an error worthy of noting specially, so that particular error
3528 62. When Exim is trying to find all the local interfaces, it used to panic and
3529 die if the ioctl to get the interface flags failed. However, it seems that
3530 on at least one OS (Solaris 9) it is possible to have an interface that is
3531 included in the list of interfaces, but for which you get a failure error
3532 for this call. This happens when the interface is not "plumbed" into a
3533 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3534 failure of the "get flags" call assumes that the interface is down.
3536 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3537 makes life easier for people who are doing arithmetic on fields extracted
3538 from dates, where you often get leading zeros that should not be
3539 interpreted as octal.
3541 64. Added qualify_domain to the redirect router, to override the global
3544 65. If a pathologically long header line contained very many addresses (the
3545 report of this problem mentioned 10 000) and each of them was rewritten,
3546 Exim could use up a very large amount of memory. (It kept on making new
3547 copies of the header line as it rewrote, and never released the old ones.)
3548 At the expense of a bit more processing, the header rewriting function has
3549 been changed so that it no longer eats memory in this way.
3551 66. The generation of the Received: header has been moved from the time that a
3552 message starts to be received, to the time that it finishes. The timestamp
3553 in the Received: header should now be very close to that of the <= log
3554 line. There are two side-effects of this change:
3556 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3557 logged header lines no longer include the local Received: line, because
3558 it has not yet been created. The same applies to a copy of the message
3559 that is returned to a non-SMTP sender when a message is rejected.
3561 (b) When a filter file is tested using -bf, no additional Received: header
3562 is added to the test message. After some thought, I decided that this
3565 This change does not affect the value of $received_for. It is still set
3566 after address rewriting, but before local_scan() is called.
3568 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3570 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3571 gave an unhelpful panic error message, and a defer error. I have managed to
3572 change this behaviour so that it now rejects any supplied certificate,
3573 which seems right, as the list of acceptable certificates is empty.
3575 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3576 gave an unhelpful defer error. I have not managed to make this reject any
3577 supplied certificates, but the error message it gives is "no certificate
3578 supplied", which is not helpful.
3580 70. exigrep's output now also includes lines that are not associated with any
3581 message, but which match the given pattern. Implemented by a patch from
3582 Martin Sluka, which also tidied up the Perl a bit.
3584 71. Recipient callout verification, like sender verification, was using <> in
3585 the MAIL FROM command. This isn't really the right thing, since the actual
3586 sender may affect whether the remote host accepts the recipient or not. I
3587 have changed it to use the actual sender in the callout; this means that
3588 the cache record is now keyed on a recipient/sender pair, not just the
3589 recipient address. There doesn't seem to be a real danger of callout loops,
3590 since a callout by the remote host to check the sender would use <>.
3591 [SEE ABOVE: changed after hitting problems.]
3593 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3594 temporary errors. However, in the case of such a code being given after
3595 the end of a data transmission (i.e. after ".") Exim was failing to write
3596 a retry record for the message. (Yes, there was some broken host that was
3597 actually sending 8xx at this point.)
3599 73. An unknown lookup type in a host list could cause Exim to panic-die when
3600 the list was checked. (An example that provoked this was putting <; in the
3601 middle of a list instead of at the start.) If this happened during a DATA
3602 ACL check, a -D file could be left lying around. This kind of configuration
3603 error no longer causes Exim to die; instead it causes a defer errror. The
3604 incident is still logged to the main and panic logs.
3606 74. Buglet left over from Exim 3 conversion. The message "too many messages
3607 in one connection" was written to the rejectlog but not the mainlog, except
3608 when address rewriting (yes!) was being logged.
3610 75. Added write_rejectlog option.
3612 76. When a system filter was run not as root (that is, when system_filter_user
3613 was set), the values of the $n variables were not being returned to the
3614 main process; thus, they were not subsequently available in the $sn
3617 77. Added +return_path_on_delivery log selector.
3619 78. A connection timeout was being treated differently from recipients deferred
3620 when testing hosts_max_try with a message that was older than the host's
3621 retry timeout. (The host should not be counted, thus allowing all hosts to
3622 be tried at least once before bouncing.) This may have been the cause of an
3623 occasionally reported bug whereby a message would remain on the queue
3624 longer than the retry timeout, but would be bounced if a delivery was
3625 forced. I say "may" because I never totally pinned down the problem;
3626 setting up timeout/retry tests is difficult. See also the next item.
3628 79. The ultimate address timeout was not being applied to errors that involved
3629 a combination of host plus message (for example, a timeout on a MAIL
3630 command). When an address resolved to a number of possible hosts, and they
3631 were not all tried for each delivery (e.g. because of hosts_max_try), a
3632 message could remain on the queue longer than the retry timeout.
3634 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3637 81. Fixed an obscure SMTP outgoing bug which required at least the following
3638 conditions: (a) there was another message waiting for the same server;
3639 (b) the server returned 5xx to all RCPT commands in the first message so
3640 that the message was not completed; (c) the server dropped the connection
3641 or gave a negative response to the RSET that Exim sends to abort the
3642 transaction. The observed case was a dropped connection after DATA that had
3643 been sent in pipelining mode. That is, the server had advertised PIPELINING
3644 but was not implementing it correctly. The effect of the bug was incorrect
3645 behaviour, such as trying another host, and this could lead to a crash.
3651 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3652 and daemon.c were passed as pointers to ints; they should have been
3653 pointers to socklen_t variables (which are typically unsigned ints).
3655 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3658 3. Fixed a really odd bug that affected only the testing scheme; patching a
3659 certain fixed string in the binary changed the value of another string that
3660 happened to be identical to the end of the original first string.
3662 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3663 name", it returns that address as the IP address. On some operating
3664 systems (e.g. Solaris), it also passes back the IP address string as the
3665 "host name". However, on others (e.g. Linux), it passes back an empty
3666 string. Exim wasn't checking for this, and was changing the host name to an
3667 empty string, assuming it had been canonicized.
3669 5. Although rare, it is permitted to have more than one PTR record for a given
3670 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3671 all the names associated with an address, because they do in Solaris.
3672 However, it seems that they do not in Linux for data that comes from the
3673 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3674 I found this out when I moved to a new Linux workstation and tried to run
3675 the Exim test suite.
3677 To get round this problem I have changed the code so that it now does its
3678 own call to the DNS to look up PTR records when searching for a host name.
3679 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3680 addresses that are only in /etc/hosts are still found.
3682 This behaviour is, however, controlled by an option called host_lookup_
3683 order, which defaults to "bydns:byaddr". If people want to use the other
3684 order, or indeed, just use one or the other means of lookup, they can
3685 specify it in this variable.
3687 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3688 some operating systems, this comes back from gethostbyaddr() as an empty
3689 string, and this is what Exim used to test for. However, it seems that in
3690 other systems, "." is yielded. Exim now tests for this case too.
3692 7. The values of check_spool_space and check_log_space are now held internally
3693 as a number of kilobytes instead of an absolute number of bytes. If a
3694 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3695 kilobyte. This means that much larger values can be stored.
3697 8. Exim monitor: an attempt to get the action menu when not actually pointing
3698 at a message produces an empty menu entitled "No message selected". This
3699 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3700 no entries in it ("Shell widget menu has zero width and/or height"). So I
3701 have added a single, blank menu entry in this case.
3703 9. Added ${quote_local_part.
3705 10. MIME decoding is now applied to the contents of Subject: header lines when
3708 11. Now that a reference to $sender_host_address automatically causes a reverse
3709 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3710 host lookup before query-style lookups in lists that might use this
3711 variable. This has therefore been abolished, and the "net-" prefix is no
3712 longer necessary for query-style lookups.
3714 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3715 have been a typo for LFLAGS, so it has been changed.
3717 13. The install script calls Exim with "-C /dev/null" in order to find the
3718 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3719 to be output. Howeve, since Exim outputs its version number before the
3720 error, it didn't break the script. It just looked ugly. I fixed this by
3721 always allowing "-C /dev/null" if the caller is root.
3723 14. Ignore overlarge ACL variable number when reading spool file - insurance
3724 against a later release with more variables having written the file.
3726 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3727 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3728 this, as well as the form without the "IPv6" on the front.
3730 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3731 OS/Makefile-Darwin file.
3733 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3734 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3736 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3737 variable. However, some OS use ints for inodes. I've added cast to long int
3738 to get rid of the compiler warning.
3740 19. I had forgotten to lock out "/../" in configuration file names when
3741 ALT_CONFIG_PREFIX was set.
3743 20. Routers used for verification do not need to specify transports. However,
3744 if such a router generated a host list, and callout was configured, Exim
3745 crashed, because it could not find a port number from the (non-existent)
3746 transport. It now assumes port 25 in this circumstance.
3748 21. Added the -t option to exigrep.
3750 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3751 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3752 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3753 them is set, LOOKUP_LSEARCH is forced.
3755 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3757 24. Added sender and host information to the "rejected by local_scan()" log
3758 line; previously there was no indication of these.
3760 25. Added .include_if_exists.
3762 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3763 Linux. It turns out that not all file systems support this. Apparently some
3764 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3765 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3766 not supported on the file descriptor, is now ignored when Exim is trying to
3767 sync a directory. This applies only to Linux.
3769 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3771 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3772 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3775 29. Added support for maildirsize files from supplied patch (modified a bit).
3777 30. The use of :fail: followed by an empty string could lead Exim to respond to
3778 sender verification failures with (e.g.):
3780 550 Verification failed for <xxx>
3781 550 Sender verify failed
3783 where the first response line was missing the '-' that indicates it is not
3784 the final line of the response.
3786 31. The loop for finding the name of the user that called Exim had a hardwired
3787 limit of 10; it now uses the value of finduser_retries, which is used for
3788 all other user lookups.
3790 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3793 33. Exim was neglecting to zero errno before one call of strtol() when
3794 expanding a string and expecting an integer value. On some systems this
3795 resulted in spurious "integer overflow" errors. Also, it was casting the
3796 result into an int without checking.
3798 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3799 did not work. The code looks as if it has *never* worked, though it appears
3800 to have been documented since at least releast 1.62. I have made it work.
3802 35. The "timeout_DNS" error in retry rules, also documented since at least
3803 1.62, also never worked. As it isn't clear exactly what this means, and
3804 clearly it isn't a major issue, I have abolished the feature by treating it
3805 as "timeout", and writing a warning to the main and panic logs.
3807 36. The display of retry rules for -brt wasn't always showing the error code
3810 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3811 timeout_connect_A, timeout_connect_MX.
3813 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3814 to the empty sender.
3816 39. The daemon was not analysing the content of -oX till after it had closed
3817 stderr and disconnected from the controlling terminal. This meant that any
3818 syntax errors were only noted on the panic log, and the return code from
3819 the command was 0. By re-arranging the code a little, I've made the
3820 decoding happen first, so such errors now appear on stderr, and the return
3821 code is 1. However, the actual setting up of the sockets still happens in
3822 the disconnected process, so errors there are still only recorded on the
3825 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3826 connections (as happens on some IP stacks) was logged at start up time as
3827 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3828 it from "IPv6 and IPv4", which means that two separate sockets are being
3831 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3832 says whether AF_INET or AF_INET6 was passed as an argument.
3834 42. Exiwhat output was messed up when time zones were included in log
3837 43. Exiwhat now gives more information about the daemon's listening ports,
3838 and whether -tls-on-connect was used.
3840 44. The "port" option of the smtp transport is now expanded.
3842 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3843 silently ignored. Now an error message is written to the main and panic
3846 46. There's a new ACL modifier called "logwrite" which writes to a log file
3847 as soon as it is encountered.
3849 47. Added $local_user_uid and $local_user_gid at routing time.
3851 48. Exim crashed when trying to verify a sender address that was being
3854 49. Exim was recognizing only a space character after ".include". It now also
3855 recognizes a tab character.
3857 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3858 extracting the relevant information from the specification. The man page no
3859 longer contains scrambled data for the -d option, and I've added a section
3860 at the front about calling Exim under different names.
3862 51. Added "extra_headers" argument to the "mail" command in filter files.
3864 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3867 53. Installed eximstats 1.29.
3869 54. Added transport_filter_timeout as a generic transport option.
3871 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3872 Cc: header lines. This was required by RFC 822, but it not required by RFC
3875 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3876 incoming messages that did not have them. Now it does so only if the
3877 message originates locally, that is, if there is no associated remote host
3878 address. When Resent- header lines are present, this applies to the Resent-
3879 lines rather than the non-Resent- lines.
3881 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3882 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3884 58. Messages for configuration errors now include the name of the main
3885 configuration file - useful now that there may be more than one file in a
3886 list (.included file names were always shown).
3888 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3889 for those rare installations that do not start the daemon as root or run it
3890 setuid root. I've cut out the call to initgroups() if the daemon is not
3893 60. The Exim user and group can now be bound into the binary as text strings
3894 that are looked up at the start of Exim's processing.
3896 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3898 62. Added $mailstore_basename variable.
3900 63. Installed patch to sieve.c from Michael Haardt.
3902 64. When Exim failed to open the panic log after failing to open the main log,
3903 the original message it was trying to log was written to stderr and debug
3904 output, but if they were not available (the usual case in production), it
3905 was lost. Now it is written to syslog before the two lines that record the
3906 failures to open the logs.
3908 65. Users' Exim filters run in subprocesses under the user's uid. It is
3909 possible for a "deliver" command or an alias in a "personal" command to
3910 provoke an address rewrite. If logging of address rewriting is configured,
3911 this fails because the process is not running as root or exim. There may be
3912 a better way of dealing with this, but for the moment (because 4.30 needs
3913 to be released), I have disabled address rewrite logging when running a
3914 filter in a non-root, non-exim process.
3920 1. The buildconfig auxiliary program wasn't quoting the value set for
3921 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3922 not defined. This bug was masked in 4.22 by the effect that was fixed in
3925 2. Some messages that were rejected after a message id was allocated were
3926 shown as "incomplete" by exigrep. It no longer does this for messages that
3927 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3929 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3930 have allow_domain_literals set, the ID did not get logged in the <= line.
3931 Domain literals are now always recognized in Message-ID: header lines.
3933 4. The first argument for a ${extract expansion item is the key name or field
3934 number. Leading and trailing spaces in this item were not being ignored,
3935 causing some misleading effects.
3937 5. When deliver_drop_privilege was set, single queue runner processes started
3938 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3939 same command in the process it spins off) were not dropping privilege.
3941 6. When the daemon running as "exim" started a queue runner, it always
3942 re-executed Exim in the spun-off process. This is a waste of effort when
3943 deliver_drop_privilege is set. The new process now just calls the
3944 queue-runner function directly.
3950 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3953 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3954 ignored. Though the use of -oP was forcing the writing of a pid file, it
3955 was always written to the default place.
3957 3. If the message "no IP address found for host xxxx" is generated during
3958 incoming verification, it is now followed by identification of the incoming
3959 connection (so you can more easily find what provoked it).
3961 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3963 5. Added some features to "harden" Exim a bit more against certain attacks:
3965 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3966 be put in Local/Makefile. This is like the never_users runtime option,
3967 but it cannot be overridden. The default setting is "root".
3969 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3970 prefix string with which any file named in a -C command line option
3973 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3974 is retained for -C and -D only if the caller of Exim is root. Without
3975 it, the exim user may also use -C and -D and retain privilege.
3977 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3978 command line option is disabled.
3980 6. Macro names set by the -D option must start with an upper case letter, just
3981 like macro names defined in the configuration file.
3983 7. Added "dereference=" facility to LDAP.
3985 8. Two instances of the typo "uknown" in the source files are fixed.
3987 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3988 the Configure-Makefile script screwed up while processing it.
3990 10. Incorporated PCRE 4.4.
3992 11. The SMTP synchronization check was not operating right at the start of an
3993 SMTP session. For example, it could not catch a HELO sent before the client
3994 waited for the greeting. There is now a check for outstanding input at the
3995 point when the greeting is written. Because of the duplex, asynchronous
3996 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3997 way, but not yet received, when the check is performed.
3999 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4000 on TCP/IP sockets, because this apparently causes some broken clients to
4003 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4004 unchanged) from the Cygwin maintainer.
4006 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4007 was supported instead of testing for mbx. Effectively a typo.
4009 15. The spa authenticator server code was not checking that the input it
4010 received was valid base64.
4012 16. The debug output line for the "set" modifier in ACLs was not showing the
4013 name of the variable that was being set.
4015 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4017 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4018 lookup on its own. Something else was needed to trigger the lookup. For
4019 example, a match in host_lookup or the need for a host name in a host list.
4020 Now, if $sender_host_name is referenced and the host name has not yet been
4021 looked up, a lookup is performed. If the lookup fails, the variable remains
4022 empty, and $host_lookup_failed is set to "1".
4024 19. Added "eqi" as a case-independent comparison operator.
4026 20. The saslauthd authentication condition could segfault if neither service
4027 nor realm was specified.
4029 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4030 error message that was logged was misleading, and incoming SMTP
4031 connections were dropped. The message is now more accurate, and temporary
4032 errors are given to SMTP connections.
4034 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4035 (except RSET) with a 421 error, until QUIT is received. However, it was
4036 failing to send a response to QUIT.
4038 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4039 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4040 ACL is now run after the helo_try_verify_hosts code.
4042 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4043 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4044 case-independent, so other case variants are also recognized). Apparently
4045 some systems use these upper case variants.
4047 25. If more than two messages were waiting for the same host, and a transport
4048 filter was specified for the transport, Exim sent two messages over the
4049 same TCP/IP connection, and then failed with "socket operation on non-
4050 socket" when it tried to send the third.
4052 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4054 27. The extern definition of crypt16() in expand.c was not being excluded when
4055 the OS had its own crypt16() function.
4057 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4058 as a preferred synonym for return_size_limit, both as an option and as an
4061 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4063 30. Changed the default configuration ACL to relax the local part checking rule
4064 for addresses that are not in any local domains. For these addresses,
4065 slashes and pipe symbols are allowed within local parts, but the sequence
4066 /../ is explicitly forbidden.
4068 31. SPA server authentication was not clearing the challenge buffer before
4071 32. log_message in a "warn" ACL statement was writing to the reject log as
4072 well as to the main log, which contradicts the documentation and doesn't
4073 seem right (because no rejection is happening). So I have stopped it.
4075 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4076 However, I am unable to do any testing of this.
4078 34. Fixed an infelicity in the appendfile transport. When checking directories
4079 for a mailbox, to see if any needed to be created, it was accidentally
4080 using path names with one or more superfluous leading slashes; tracing
4081 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4083 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4084 contents are added to the log line that is written for every discarded
4085 recipient. (Previously a log_message setting was ignored.)
4087 36. The ${quote: operator now quotes the string if it is empty.
4089 37. The install script runs exim in order to find its version number. If for
4090 some reason other than non-existence or emptiness, which it checks, it
4091 could not run './exim', it was installing it with an empty version number,
4092 i.e. as "exim-". This error state is now caught, and the installation is
4095 38. An argument was missing from the function that creates an error message
4096 when Exim fails to connect to the socket for saslauthd authentication.
4097 This could cause Exim to crash, or give a corrupted message.
4099 39. Added isip, isip4, and isip6 to ${if conditions.
4101 40. The ACL variables $acl_xx are now saved with the message, and can be
4102 accessed later in routers, transports, and filters.
4104 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4105 strings in the file are not string-expanded.
4107 42. If a MAIL command specified a SIZE value that was too large to fit into an
4108 int variable, the check against message_size_limit failed. Such values are
4109 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4110 day this will have to be increased, but I don't think I want to be around
4111 when emails are that large.
4118 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4119 iconv() is not standard in FreeBSD.
4121 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4122 IPv6 enabled. The observed symptom was a segmentation fault on return from
4123 the function os_common_find_running_interfaces() in src/os.c.
4125 3. In the check_special_case() function in daemon.c I had used "errno" as an
4126 argument name, which causes warnings on some systems. This was basically a
4127 typo, since it was named "eno" in the comments!
4129 4. The code that waits for the clock to tick (at a resolution of some fraction
4130 of a second) so as to ensure message-id uniqueness was always waiting for
4131 at least one whole tick, when it could have waited for less. [This is
4132 almost certainly not relevant at current processor speeds, where it is
4133 unlikely to ever wait at all. But we try to future-proof.]
4135 5. The function that sleeps for a time interval that includes fractions of a
4136 second contained a race. It did not block SIGALRM between setting the
4137 timer, and suspending (a couple of lines later). If the interval was short
4138 and the sigsuspend() was delayed until after it had expired, the suspension
4139 never ended. On busy systems this could lead to processes getting stuck for
4142 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4143 process, before it forks any delivery processes. The open lookup caching
4144 mechanism meant that the open file or database connection was passed into
4145 the delivery process. The problem was that delivery processes always tidy
4146 up cached lookup data. This could cause a problem for the next delivery
4147 process started by the queue runner, because the external queue runner
4148 process does not know about the closure. So the next delivery process
4149 still has data in the lookup cache. In the case of a file lookup, there was
4150 no problem because closing a file descriptor in a subprocess doesn't affect
4151 the parent. However, if the lookup was caching a connection to a database,
4152 the connection was closed, and the second delivery process was likely to
4153 see errors such as "PGSQL: query failed: server closed the connection
4154 unexpectedly". The problem has been fixed by closing all cached lookups
4155 in a queue runner before running a delivery process.
4157 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4158 seem to have the "const" qualifier which it has on other OS. I've
4161 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4162 *of the same type* (client or server) with the same public name that an
4163 error should be diagnosed.
4165 9. When Exim looked up a host name for an IP address, but failed to find the
4166 original IP address when looking up the host name (a safety check), it
4167 output the message "<ip address> does not match any IP for NULL", which was
4168 confusing, to say the least. The bug was that the host name should have
4169 appeared instead of "NULL".
4171 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4172 user that is built into the binary, and the -C or -D options is used, root
4173 privilege is dropped before the configuration file is read. In addition,
4174 logging is switched to stderr instead of the normal log files. If the
4175 configuration then re-defines the Exim user, the unprivileged environment
4176 is probably not what is expected, so Exim logs a panic warning message (but
4179 However, if deliver_drop_privilege is set, the unprivileged state may well
4180 be exactly what is intended, so the warning has been cut out in that case,
4181 and Exim is allowed to try to write to its normal log files.
4187 1. smtp_return_error_details was not giving details for temporary sender
4188 or receiver verification errors.
4190 2. Diagnose a configuration error if two authenticators have the same public
4193 3. Exim used not to create the message log file for a message until the first
4194 delivery attempt. This could be confusing when incoming messages were held
4195 for policy or load reasons. The message log file is now created at the time
4196 the message is received, and an initial "Received" line is written to it.
4198 4. The automatically generated man page for command line options had a minor
4199 bug that caused no ill effects; however, a more serious problem was that
4200 the procedure for building the man page automatically didn't always
4201 operate. Consequently, release 4.20 contains an out-of-date version. This
4202 shouldn't happen again.
4204 5. When building Exim with embedded Perl support, the script that builds the
4205 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4206 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4208 6. The freeze_tell option was not being used for messages that were frozen on
4209 arrival, either by an ACL or by local_scan().
4211 7. Added the smtp_incomplete_transaction log selector.
4213 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4214 was accepting AUTH without a new EHLO.
4216 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4217 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4220 10. Logging of TCP/IP connections (when configured) now happens in the main
4221 daemon process instead of the child process, so that the TCP/IP connection
4222 count is more accurate (but it can never be perfect).
4224 11. The use of "drop" in a nested ACL was not being handled correctly in the
4225 outer ACL. Now, if condition failure induced by the nested "drop" causes
4226 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4227 or "require"), the connection is dropped.
4229 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4230 that yield "discard" can now be used with an "accept" or a "discard" verb,
4231 but an error is generated for any others (because I can't see a useful way
4232 to define what should happen).
4234 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4235 are now processed in the same way as lines in the Exim configuration file.
4236 In particular, continuation lines are supported.
4238 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4240 15. Added -ti meaning -t -i.
4242 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4243 domains, and warn by logging if others are found.
4245 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4246 returned from GIFCONF when Exim is trying to find the list of interfaces on
4247 a host. The code in os.c has been modified to copy each ifreq to an aligned
4248 structure in all cases.
4250 Also, in some cases, the returned ifreq's were being copied to a 'struct
4251 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4252 means the last couple of bytes of an IPv6 address could be chopped if the
4253 ifreq contained only a normal sockaddr (14 bytes storage).
4255 18. Named domain lists were not supported in the hosts_treat_as_local option.
4256 An entry such as +xxxx was not recognized, and was treated as a literal
4259 19. Ensure that header lines added by a DATA ACL are included in the reject log
4260 if the ACL subsequently rejects the message.
4262 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4263 MD5 (which is deprecated).
4265 21. When testing a filter file using -bf, Exim was writing a message when it
4266 took the sender from a "From " line in the message, but it was not doing so
4267 when it took $return_path from a Return-Path: header line. It now does.
4269 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4270 with a valid header line field name (a series of printing characters
4271 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4273 23. Changed "disc" in the source to "disk" to conform to the documentation and
4274 the book and for uniformity.
4276 24. Ignore Sendmail's -Ooption=value command line item.
4278 25. When execve() failed while trying to run a command in a pipe transport,
4279 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4280 could be confused with a return value of 69 from the command itself. This
4281 has been changed to 127, the value the shell returns if it is asked to run
4282 a non-existent command. The wording for the related log line suggests a
4283 non-existent command as the problem.
4285 26. If received_header_text expands to an empty string, do not add a Received:
4286 header line to the message. (Well, it adds a token one on the spool, but
4287 marks it "old" so that it doesn't get used or transmitted.)
4289 27. Installed eximstats 1.28 (addition of -nt option).
4291 28. There was no check for failure on the call to getsockname() in the daemon
4292 code. This can fail if there is a shortage of resources on the system, with
4293 ENOMEM, for example. A temporary error is now given on failure.
4295 29. Contrary to the C standard, it seems that in some environments, the
4296 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4297 program. Exim now does this explicitly; it affects the formatting of
4298 timestamps using strftime().
4300 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4301 complaints. I've now initialized all the variables, to avoid this.
4303 32. Header lines added by a system filter were not being "seen" during
4304 transport-time rewrites.
4306 33. The info_callback() function passed to OpenSSL is set up with type void
4307 (*)(SSL *, int, int), as described somewhere. However, when calling the
4308 function (actually a macro) that sets it up, the type void(*)() is
4309 expected. I've put in a cast to prevent warnings from picky compilers.
4311 34. If a DNS black list lookup found a CNAME record, but there were no A
4312 records associated with the domain it pointed at, Exim crashed.
4314 35. If a DNS black list lookup returned more than one A record, Exim ignored
4315 all but the first. It now scans all returned addresses if a particular IP
4316 value is being sought. In this situation, the contents of the
4317 $dnslist_value variable are a list of all the addresses, separated by a
4320 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4321 to accept a host name and all its aliases if the forward lookup for any of
4322 them yielded the IP address of the incoming connection. Now it accepts only
4323 those names whose forward lookup yields the correct IP address. Any other
4324 names are discarded. This closes a loophole whereby a rogue DNS
4325 administrator could create reverse DNS records to break through a
4326 wildcarded host restriction in an ACL.
4328 37. If a user filter or a system filter that ran in a subprocess used any of
4329 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4330 the wrong values were passed to the pipe command ($thisaddress had the
4331 value of $0, $0 had the value of $1, etc). This bug was introduced by
4332 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4334 38. Improved the line breaking for long SMTP error messages from ACLs.
4335 Previously, if there was no break point between 40 and 75 characters, Exim
4336 left the rest of the message alone. Two changes have been made: (a) I've
4337 reduced the minimum length to 35 characters; (b) if it can't find a break
4338 point between 35 and 75 characters, it looks ahead and uses the first one
4339 that it finds. This may give the occasional overlong line, but at least the
4340 remaining text gets split now.
4342 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4343 file descriptors might be low, and that setting 1000 would always raise it.
4344 It turns out that in some environments, the limit is already over 1000 and
4345 that lowering it causes trouble. So now Exim takes care not to decrease it.
4347 40. When delivering a message, the value of $return_path is set to $sender_
4348 address at the start of routing (routers may change the value). By an
4349 oversight, this default was not being set up when an address was tested by
4350 -bt or -bv, which affected the outcome if any router or filter referred to
4353 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4354 log only when "message" or "log_message" are set. However, if one of the
4355 conditions was an address verification, or a call to a nested ACL, the
4356 messages generated by the underlying test were being passed through. This
4357 no longer happens. The underlying message is available in $acl_verify_
4358 message for both "message" and "log_message" expansions, so it can be
4359 passed through if needed.
4361 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4362 new expansion $bh_ to give the encoded byte string without charset
4363 translation. Translation happens only if iconv() is available; HAVE_ICONV
4364 indicates this at build time. HEADERS_CHARSET gives the charset to
4365 translate to; headers_charset can change it in the configuration, and
4366 "headers charset" can change it in an individual filter file.
4368 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4369 that creates RFC 2047 encoded "words" labels them as that charset instead
4370 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4371 expansion operator; (ii) when Exim creates a From: line for a local
4372 message; (iii) when a header line is rewritten to include a "phrase" part.
4374 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4375 buggy, causing it to skip the first lines of messages whose message ID
4376 ended in 'D'. This would not have bitten before Exim release 4.14, because
4377 message IDs were unlikely to end in 'D' before then. The effect was to have
4378 incorrect size information for certain domains.
4380 45. #include "config.h" was missing at the start of the crypt16.c module. This
4381 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4384 46. If there was a timeout during a "random" callout check, Exim treated it as
4385 a failure of the random address, and carried on sending RSET and the real
4386 address. If the delay was just some slowness somewhere, the response to the
4387 original RCPT would be taken as a response to RSET and so on, causing
4388 mayhem of various kinds.
4390 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4391 when I implemented it. It didn't allow for the fact that some option values
4392 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4393 the right test for positive values.
4395 48. Domain names in DNS records are case-independent. Exim always looks them up
4396 in lower case. Some resolvers return domain names in exactly the case they
4397 appear in the zone file, that is, they may contain uppercase letters. Not
4398 all resolvers do this - some return always lower case. Exim was treating a
4399 change of case by a resolver as a change of domain, similar to a widening
4400 of a domain abbreviation. This triggered its re-routing code and so it was
4401 trying to route what was effectively the same domain again. This normally
4402 caused routing to fail (because the router wouldn't handle the domain
4403 twice). Now Exim checks for this case specially, and just changes the
4404 casing of the domain that it ultimately uses when it transmits the message
4407 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4410 50. If a filter generated a file delivery with a non-absolute name (possible if
4411 no home directory exists for the router), the forbid_file option was not
4414 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4415 the existing equality matching.
4417 52. Exim was using ints instead of ino_t variables in some places where it was
4418 dealing with inode numbers.
4420 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4421 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4422 called TMPDIR, and if it finds it is different, it changes its value.
4424 54. The smtp_printf() function is now made available to local_scan() so
4425 additional output lines can be written before returning. There is also an
4426 smtp_fflush() function to enable the detection of a dropped connection.
4427 The variables smtp_input and smtp_batched_input are exported to
4430 55. Changed the default runtime configuration: the message "Unknown user"
4431 has been removed from the ACL, and instead placed on the localuser router,
4432 using the cannot_route_message feature. This means that any verification
4433 failures that generate their own messages won't get overridden. Similarly,
4434 the "Unrouteable address" message that was in the ACL for unverifiable
4435 relay addresses has also been removed.
4437 56. Added hosts_avoid_esmtp to the smtp transport.
4439 57. The exicyclog script was not checking for the esoteric option
4440 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4441 will work only if exicyclog is run under the appropriate euid.
4443 58. Following a discussion on the list, the rules by which Exim recognises line
4444 endings on incoming messages have been changed. The -dropcr and drop_cr
4445 options are now no-ops, retained only for backwards compatibility. The
4446 following line terminators are recognized: LF CRLF CR. However, special
4447 processing applies to CR:
4449 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4450 nor a local message in the state where . is a terminator.
4452 (ii) If a bare CR is encountered in a header line, an extra space is added
4453 after the line terminator so as not to end the header. The reasoning
4454 behind this is that bare CRs in header lines are most likely either
4455 to be mistakes, or people trying to play silly games.
4457 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4458 was being incorrectly given as 18 bytes larger than it should have been.
4459 This is a VOB (very old bug).
4461 60. This may never have affected anything current, but just in case it has:
4462 When the local host is found other than at the start of a list of hosts,
4463 the local host, those with the same MX, and any that follow, are discarded.
4464 When the list in question was part of a longer list of hosts, the following
4465 hosts (not currently being processed) were also being discarded. This no
4466 longer happens. I'm not sure if this situation could ever has previously
4469 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4472 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4473 header. This is something that is recommended in a new Internet Draft, and
4474 is something that is documented as being done by Sendmail. There are two
4475 possible values. For messages generated by the autoreply transport, Exim
4478 Auto-Submitted: auto-replied
4480 whereas for all other generated messages (e.g. bounces) it adds
4482 Auto-Submitted: auto-generated
4484 63. The "personal" condition in filters now includes a test for the
4485 Auto-Submitted: header. If it contains the string "auto-" the message it
4486 not considered personal.
4488 64. Added rcpt_include_affixes as a generic transport option.
4490 65. Added queue_only_override (default true).
4492 66. Added the syslog_duplication option.
4494 67. If what should have been the first header line of a message consisted of
4495 a space followed by a colon, Exim was mis-interpreting it as a header line.
4496 It isn't of course - it is syntactically invalid and should therefore be
4497 treated as the start of the message body. The misbehaviour could have
4498 caused a number of strange effects, including loss of data in subsequent
4499 header lines, and spool format errors.
4501 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4502 client host had authenticated. This control can now be exercised by an ACL
4503 for more flexibility.
4505 69. By default, callouts do not happen when testing with -bh. There is now a
4506 variant, -bhc, which does actually run the callout code, including
4507 consulting and updating the callout cache.
4509 70. Added support for saslauthd authentication, courtesy of Alexander
4512 71. If statvfs() failed on the spool or log directories while checking their
4513 size for availability, Exim confusingly gave the error "space shortage".
4514 Furthermore, in debugging mode it crashed with a floating point exception.
4515 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4516 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4517 serious problem, Exim now writes to the main and panic logs when this
4518 happens, with details of the failure. It then refuses to accept the
4519 incoming message, giving the message "spool directory problem" or "log
4520 directory problem" with a 421 code for SMTP messages.
4522 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4523 0, 1, and 2 exist, because some OS complain for execs without them (see
4524 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4525 descriptors. However, the code omitted to check that the open succeeded,
4526 causing mysterious errors if for some reason the permissions on /dev/null
4527 got screwed. Now Exim writes a message to the main and panic logs, and
4528 bombs out if it can't open /dev/null.
4530 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4531 interact so that it is all more flexible. It is supposed to remain
4532 backwards compatible. Also added extra_local_interfaces.
4534 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4535 to bomb out with an assertion failure - to the client this appears as a
4536 connection drop. This problem occurs in the part of the code that was taken
4537 from the Samba project. Fortunately, the assertion is in a very simple
4538 function, so I have fixed this by reproducing the function inline in the
4539 one place where it is called, and arranging for authentication to fail
4540 instead of killing the process with assert().
4542 75. The SPA client code was not working when the server requested OEM rather
4543 than Unicode encoding.
4545 76. Added code to make require_files with a specific uid setting more usable in
4546 the case where statting the file as root fails - usually a non-root-mounted
4547 NFS file system. When this happens and the failure is EACCES, Exim now
4548 forks a subprocess and does the per-uid checking as the relevant uid.
4550 77. Added process_log_path.
4552 78. If log_file_path was not explicitly set, a setting of check_log_space or
4553 check_log_inodes was ignored.
4555 79. If a space check for the spool or log partitions fails, the incident is now
4556 logged. Of course, in the latter case the data may get lost...
4558 80. Added the %p formatting code to string_format() so that it can be used to
4559 print addresses in debug_print(). Adjusted all the address printing in the
4560 debugging in store.c to use %p rather than %d.
4562 81. There was a concern that a line of code in smtp_in.c could overflow a
4563 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4564 initially expressed, the concern was not well-founded, because trailing
4565 spaces are removed early. However, if the trailing spaces were followed by
4566 a NULL, they did not get removed, so the overflow was possible. Two fixes
4569 (a) I re-wrote the offending code in a cleaner fashion.
4570 (b) If an incoming SMTP command contains a NULL character, it is rejected
4573 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4574 runs initgroups(), so that if the Exim user is in any additional groups,
4575 they will be used during message reception.
4581 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob