1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.468 2007/02/05 12:35:46 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
63 PH/11 Added control=no_pipelining.
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
72 PH/14 Added log_selector = +pid.
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
80 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
81 fixed by 4.65/MH/01 (is this a record?) are fixed:
83 (i) An empty string was always treated as zero by the numeric comparison
84 operators. This behaviour has been restored.
86 (ii) It is documented that the numeric comparison operators always treat
87 their arguments as decimal numbers. This was broken in that numbers
88 starting with 0 were being interpreted as octal.
90 While fixing these problems I realized that there was another issue that
91 hadn't been noticed. Values of message_size_limit (both the global option
92 and the transport option) were treated as octal if they started with 0.
93 The documentation was vague. These values are now always treated as
94 decimal, and I will make that clear in the documentation.
100 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
101 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
104 MH/01 Don't check that the operands of numeric comparison operators are
105 integers when their expansion is in "skipping" mode (fixes bug
106 introduced by 4.64-PH/07).
108 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
109 child addresses, Exim now panics and dies. Previously, because the count
110 is held in a short int, deliveries were likely to be lost. As such a
111 large number of recipients for a single message is ridiculous
112 (performance will be very, very poor), I have chosen to impose a limit
113 rather than extend the field.
119 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
120 leftover -K file (the existence of which was triggered by #402).
121 While we were at it, introduced process PID as part of the -K
122 filename. This should rule out race conditions when creating
125 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
126 processing considerably. Previous code took too long for large mails,
127 triggering a timeout which in turn triggers #401.
129 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
130 in the DK code in transports.c. sendfile() is not really portable,
131 hence the _LINUX specificness.
133 TF/01 In the add_headers option to the mail command in an Exim filter,
134 there was a bug that Exim would claim a syntax error in any
135 header after the first one which had an odd number of characters
138 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
139 callout verification, Exim cached a "reject" for the entire domain. This
140 is correct for most verifications, but it is not correct for a recipient
141 verification with use_sender or use_postmaster set, because in that case
142 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
143 case of MAIL FROM:<> rejection from other early rejections (e.g.
144 rejection of HELO). When verifying a recipient using a non-null MAIL
145 address, the cache is ignored if it shows MAIL FROM:<> rejection.
146 Whatever the result of the callout, the value of the domain cache is
147 left unchanged (for any other kind of callout, getting as far as trying
148 RCPT means that the domain itself is ok).
150 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
153 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
154 manifest itself as EPIPE rather than ECONNECT. When tidying away a
155 session, the daemon ignores ECONNECT errors and logs others; it now
156 ignores EPIPE as well.
158 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
159 (quoted-printable decoding).
161 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
162 later the small subsequent patch to fix an introduced bug.
164 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
166 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
168 PH/08 An error is now given if message_size_limit is specified negative.
170 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
171 to be given (somewhat) arbitrary names.
173 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
176 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
177 miscellaneous code fixes
179 PH/10 Added the log_reject_target ACL modifier to specify where to log
182 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
183 hostname. This is wrong, because it relates to the incoming message (and
184 probably the interface on which it is arriving) and not to the outgoing
185 callout (which could be using a different interface). This has been
186 changed to use the value of the helo_data option from the smtp transport
187 instead - this is what is used when a message is actually being sent. If
188 there is no remote transport (possible with a router that sets up host
189 addresses), $smtp_active_hostname is used.
191 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
192 tweaks were necessary in order to get it to work (see also 21 below):
193 (a) The code assumed that strncpy() returns a negative number on buffer
194 overflow, which isn't the case. Replaced with Exim's string_format()
196 (b) There were several signed/unsigned issues. I just did the minimum
197 hacking in of casts. There is scope for a larger refactoring.
198 (c) The code used strcasecmp() which is not a standard C function.
199 Replaced with Exim's strcmpic() function.
200 (d) The code set only $1; it now sets $auth1 as well.
201 (e) A simple test gave the error "authentication client didn't specify
202 service in request". It would seem that Dovecot has changed its
203 interface. Fortunately there's a specification; I followed it and
204 changed what the client sends and it appears to be working now.
206 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
209 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
210 address A is aliased to B and C, where B exists and C does not. Without
211 -v the output is "A verified" because verification stops after a
212 successful redirection if more than one address is generated. However,
213 with -v the child addresses are also verified. Exim was outputting "A
214 failed to verify" and then showing the successful verification for C,
215 with its parentage. It now outputs "B failed to verify", showing B's
216 parentage before showing the successful verification of C.
218 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
219 look up a TXT record in a specific list after matching in a combined
222 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
223 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
224 they consult the DNS. I had assumed they would set it the way they
225 wanted; and indeed my experiments on Linux seem to show that in some
226 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
227 To be on the safe side, however, I have now made the interface to
228 host_find_byname() similar to host_find_bydns(), with an argument
229 containing the DNS resolver options. The host_find_byname() function now
230 sets these options at its start, just as host_find_bydns() does. The smtp
231 transport options dns_qualify_single and dns_search_parents are passed to
232 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
233 of host_find_byname() use the default settings of RES_DEFNAMES
234 (qualify_single) but not RES_DNSRCH (search_parents).
236 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
237 spool_read_header() do less string testing, by means of a preliminary
238 switch on the second character of optional "-foo" lines. (This is
239 overdue, caused by the large number of possibilities that now exist.
240 Originally there were few.) While I was there, I also converted the
241 str(n)cmp tests so they don't re-test the leading "-" and the first
242 character, in the hope this might squeeze out yet more improvement.
244 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
245 flag allowing group syntax was set by the header_syntax check but not
246 turned off, possible causing trouble later; (2) The flag was not being
247 set at all for the header_verify test, causing "group"-style headers to
248 be rejected. I have now set it in this case, and also caused header_
249 verify to ignore an empty address taken from a group. While doing this, I
250 came across some other cases where the code for allowing group syntax
251 while scanning a header line wasn't quite right (mostly, not resetting
252 the flag correctly in the right place). These bugs could have caused
253 trouble for malformed header lines. I hope it is now all correct.
255 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
256 with the "reply" argument non-NULL. The code, however (which originally
257 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
258 but it didn't always do it. This confused somebody who was copying the
259 code for some other use. I have removed all the tests.
261 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
262 feature that was used to support insecure browsers during the U.S. crypto
263 embargo. It requires special client support, and Exim is probably the
264 only MTA that supported it -- and would never use it because real RSA is
265 always available. This code has been removed, because it had the bad
266 effect of slowing Exim down by computing (never used) parameters for the
267 RSA_EXPORT functionality.
269 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
270 authenticator to fail if there's a tab character in the incoming data
271 (there should never be unless someone is messing about, as it's supposed
272 to be base64-encoded). Also added, on Timo's advice, the "secured" option
273 if the connection is using TLS or if the remote IP is the same as the
274 local IP, and the "valid-client-cert option" if a client certificate has
277 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
278 authenticators. This can be used for authorization after authentication
279 succeeds. (In the case of plaintext, it servers for both authentication
282 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
283 if any retry times were supplied.
285 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
286 connection, where there is no client IP address to check. In this
287 situation, the verify now always succeeds.
289 PH/25 Applied John Jetmore's -Mset patch.
291 PH/26 Added -bem to be like -Mset, but loading a message from a file.
293 PH/27 In a string expansion for a processed (not raw) header when multiple
294 headers of the same name were present, leading whitespace was being
295 removed from all of them, but trailing whitespace was being removed only
296 from the last one. Now trailing whitespace is removed from each header
297 before concatenation. Completely empty headers in a concatenation (as
300 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
301 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
303 PH/29 [Removed. This was a change that I later backed out, and forgot to
304 correct the ChangeLog entry (that I had efficiently created) before
305 committing the later change.]
307 PH/30 Exim was sometimes attempting to deliver messages that had suffered
308 address errors (4xx response to RCPT) over the same connection as other
309 messages routed to the same hosts. Such deliveries are always "forced",
310 so retry times are not inspected. This resulted in far too many retries
311 for the affected addresses. The effect occurred only when there were more
312 hosts than the hosts_max_try setting in the smtp transport when it had
313 the 4xx errors. Those hosts that it had tried were not added to the list
314 of hosts for which the message was waiting, so if all were tried, there
315 was no problem. Two fixes have been applied:
317 (i) If there are any address or message errors in an SMTP delivery, none
318 of the hosts (tried or untried) are now added to the list of hosts
319 for which the message is waiting, so the message should not be a
320 candidate for sending over the same connection that was used for a
321 successful delivery of some other message. This seems entirely
322 reasonable: after all the message is NOT "waiting for some host".
323 This is so "obvious" that I'm not sure why it wasn't done
324 previously. Hope I haven't missed anything, but it can't do any
325 harm, as the worst effect is to miss an optimization.
327 (ii) If, despite (i), such a delivery is accidentally attempted, the
328 routing retry time is respected, so at least it doesn't keep
329 hammering the server.
331 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
332 in ${readsocket because some servers need this prod.
334 PH/32 Added some extra debug output when updating a wait-xxx database.
336 PH/33 The hint "could be header name not terminated by colon", which has been
337 given for certain expansion errors for a long time, was not being given
338 for the ${if def:h_colon_omitted{... case.
340 PH/34 The spec says: "With one important exception, whenever a domain list is
341 being scanned, $domain contains the subject domain." There was at least
342 one case where this was not true.
344 PH/35 The error "getsockname() failed: connection reset by peer" was being
345 written to the panic log as well as the main log, but it isn't really
346 panic-worthy as it just means the connection died rather early on. I have
347 removed the panic log writing for the ECONNRESET error when getsockname()
350 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
351 runs only) independently of the message's sender address. This meant
352 that, if the 4xx error was in fact related to the sender, a different
353 message to the same recipient with a different sender could confuse
354 things. In particualar, this can happen when sending to a greylisting
355 server, but other circumstances could also provoke similar problems.
356 I have changed the default so that the retry time for these errors is now
357 based a combination of the sender and recipient addresses. This change
358 can be overridden by setting address_retry_include_sender=false in the
361 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
362 remote server are returned as part of bounce messages. This was not
363 happening for LMTP over a pipe (the lmtp transport), but now it is the
364 same for both kinds of LMTP.
366 PH/38 Despite being documented as not happening, Exim was rewriting addresses
367 in header lines that were in fact CNAMEs. This is no longer the case.
369 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
370 and queue runs started by the daemon processed all messages. This has
371 been fixed so that -R and -S can now usefully be given with -q<time>.
373 PH/40 Import PCRE release 6.7 (fixes some bugs).
375 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
377 PH/42 Give an error if -q is specified more than once.
379 PH/43 Renamed the variables $interface_address and $interface_port as
380 $received_ip_address and $received_port, to make it clear that these
381 values apply to message reception, and not to the outgoing interface when
382 a message is delivered. (The old names remain recognized, of course.)
384 PH/44 There was no timeout on the connect() call when using a Unix domain
385 socket in the ${readsocket expansion. There now is.
387 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
388 be meaningful with "accept".
390 SC/01 Eximstats V1.43
391 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
393 SC/02 Eximstats V1.44
394 Use a glob alias rather than an array ref in the generated
395 parser. This improves both readability and performance.
397 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
398 Collect SpamAssassin and rejection statistics.
399 Don't display local sender or destination tables unless
400 there is data to show.
401 Added average volumes into the top table text output.
403 SC/04 Eximstats V1.46
404 Collect data on the number of addresses (recipients)
405 as well as the number of messages.
407 SC/05 Eximstats V1.47
408 Added 'Message too big' to the list of mail rejection
409 reasons (thanks to Marco Gaiarin).
411 SC/06 Eximstats V1.48
412 Mainlog lines which have GMT offsets and are too short to
413 have a flag are now skipped.
415 SC/07 Eximstats V1.49 (Alain Williams)
416 Added the -emptyok flag.
418 SC/08 Eximstats V1.50
419 Fixes for obtaining the IP address from reject messages.
421 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
422 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
423 whitesspace changes from 4.64-PH/27
425 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
428 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
429 are found, allow negative numbers in numeric criteria)
431 JJ/06 exipick.20061117.2, added new $message_body_missing variable
433 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
434 to match changes made in 4.64-PH/43
436 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
438 PH/47 Put in an explicit test for a DNS lookup of an address record where the
439 "domain" is actually an IP address, and force a failure. This locks out
440 those revolvers/nameservers that support "A-for-A" lookups, in
441 contravention of the specifications.
443 PH/48 When a host name was looked up from an IP address, and the subsequent
444 forward lookup of the name timed out, the host name was left in
445 $sender_host_name, contrary to the specification.
447 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
448 restricted to single-key lookups, Exim was not diagnosing an error if
449 * or *@ was used with a query-style lookup.
451 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
453 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
454 long ago, but noone interested enough thought of it. Let's just say that
455 the "1.1" means that there are some new functions that weren't there at
456 some point in the past.
458 PH/51 Error processing for expansion failure of helo_data from an smtp
459 transport during callout processing was broken.
461 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
462 tested/used via the -bh/-bhc/-bs options.
464 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
465 bug, fixed in subsequent PCRE releases).
467 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
468 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
470 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
476 SC/01 Use a glob alias rather than an array ref in eximstats generated
477 parser. This improves both readability and performance.
479 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
480 Don't display local sender or destination tables in eximstats unless
481 there is data to show.
482 Added average volumes into the eximstats top table text output.
484 SC/03 Collect data on the number of addresses (recipients) as well
485 as the number of messages in eximstats.
487 TF/01 Correct an error in the documentation for the redirect router. Exim
488 does (usually) call initgroups() when daemonizing.
490 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
491 with consistent privilege compared to when running as a daemon.
493 TF/03 Note in the spec that $authenticated_id is not set for local
494 submissions from trusted users.
496 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
497 Thanks to Dean Brooks <dean@iglou.com> for the patch.
499 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
500 by adding some example configuration directives to the default
501 configuration file. A little bit of work is required to uncomment the
502 directives and define how usernames and passwords are checked, but
503 there is now a framework to start from.
505 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
506 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
507 without this. I don't know how relevant this is to other LDAP libraries.
509 PH/02 Add the verb name to the "unknown ACL verb" error.
511 PH/03 Magnus Holmgren's patch for filter_prepend_home.
513 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
515 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
516 directory not expanded when it should be if an expanded home directory
517 was set for the address (which is overridden by the transport).
519 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
522 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
523 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
524 because it is too late at that time, and has no effect.
526 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
527 security issue with \' (bugzilla #107). I could not use the
528 PQescapeStringConn() function, because it needs a PGconn value as one of
531 PH/08 When testing addresses using -bt, indicate those final addresses that
532 are duplicates that would not cause an additional delivery. At least one
533 person was confused, thinking that -bt output corresponded to deliveries.
534 (Suppressing duplicates isn't a good idea as you lose the information
535 about possibly different redirections that led to the duplicates.)
537 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
538 systems where poll() doesn't work, in particular OS X.
540 PH/10 Added more information to debugging output for retry time not reached.
542 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
543 operations in malware.c.
545 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
548 PH/13 If write_rejectlog was set false when logging was sent to syslog with
549 syslog_duplication set false, log lines that would normally be written
550 both the the main log and to the reject log were not written to syslog at
553 PH/14 In the default configuration, change the use of "message" in ACL warn
554 statements to "add_header".
556 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
557 not followed by a command (e.g. "seen endif").
559 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
560 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
563 PH/17 Added extra conditions to the default value of delay_warning_condition
567 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
568 { match{$h_precedence:}{(?i)bulk|list|junk} } \
569 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
572 The Auto-Submitted: and various List- headers are standardised, whereas I
573 don't think Precedence: ever was.
575 PH/18 Refactored debugging code in route_finduser() to show more information,
576 in particular, the error code if getpwnam() issues one.
578 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
579 This is apparently needed in addition to the PH/07 change above to avoid
580 any possible encoding problems.
582 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
583 but not after initializing Perl.
585 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
586 output them only if debugging. By default they are written stderr,
587 apparently, which is not desirable.
589 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
592 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
595 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
597 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
598 authenticated or an ident call has been made. Suppress the default
599 values for $authenticated_id and $authenticated_sender (but permit -oMai
600 and -oMas) when testing with -bh.
602 PH/24 Re-jigged the order of the tests in the default configuration so that the
603 tests for valid domains and recipients precede the DNS black list and CSA
604 tests, on the grounds that those ones are more expensive.
606 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
607 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
608 This bug exists in every version of Exim that I still have, right back to
611 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
612 However, an attempt to turn on case-sensitivity in a regex key by
613 including (?-i) didn't work because the subject string was already
614 lowercased, and the effects were non-intuitive. It turns out that a
615 one-line patch can be used to allow (?-i) to work as expected.
621 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
622 other effects) broke the use of negated acl sub-conditions.
624 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
627 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
628 "Deny" causes Exim to reject the incoming connection with a 554 error.
629 Unfortunately, if there is a major crisis, such as a disk failure,
630 tcp-wrappers gives "deny", whereas what one would like would be some
631 kind of temporary error. A kludge has been added to help with this.
632 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
633 554 error is used if errno is still zero or contains ENOENT (which occurs
634 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
637 PH/03 Add -lutil to the default FreeBSD LIBS setting.
639 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
640 errors. Otherwise a message that provokes a temporary error (when other
641 messages do not) can cause a whole host to time out.
643 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
644 addresses were routed directly to files or pipes from a redirect router.
645 File deliveries just didn't batch; pipe deliveries might have suffered
648 PH/06 A failure to get a lock for a hints database would erroneously always say
649 "Failed to get write lock", even when it was really a read lock.
651 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
652 of 0600. This has been changed to use the value of the lockfile_mode
653 option (which defaults to 0600).
655 PH/08 Applied small patch from the Sieve maintainer.
657 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
658 folder from quota calculations, a direct delivery into this folder messed
659 up the contents of the maildirsize file. This was because the regex was
660 used only to exclude .Trash (or whatever) when the size of the mailbox
661 was calculated. There was no check that a delivery was happening into an
662 excluded directory. This bug has been fixed by ignoring all quota
663 processing for deliveries into excluded directories.
665 PH/10 Added the maildirfolder_create_regex option to appendfile.
671 PH/01 The code for finding all the local interface addresses on a FreeBSD
672 system running IPv6 was broken. This may well have applied to all BSD
673 systems, as well as to others that have similar system calls. The broken
674 code found IPv4 interfaces correctly, but gave incorrect values for the
675 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
676 that it would not match correctly against @[] and not recognize the IPv6
679 PH/02 The ipliteral router was not recognizing addresses of the form user@
680 [ipv6:....] because it didn't know about the "ipv6:" prefix.
682 PH/03 Added disable_ipv6.
684 PH/04 Changed $reply_address to use the raw form of the headers instead of the
685 decoded form, because it is most often used to construct To: headers
686 lines in autoreplies, and the decoded form may well be syntactically
687 invalid. However, $reply_address has leading white space removed, and all
688 newlines turned into spaces so that the autoreply transport does not
691 PH/05 If group was specified without a user on a router, and no group or user
692 was specified on a transport, the group from the router was ignored.
694 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
695 for visible compile-time settings that can be used to change these
696 numbers, for those that want even more. Backwards compatibility with old
697 spool files has been maintained. However, going back to a previous Exim
698 release will lost any variables that are in spool files.
700 PH/07 Two small changes when running in the test harness: increase delay when
701 passing a TCP/IP connection to a new process, in case the original
702 process has to generate a bounce, and remove special handling of
703 127.0.0.2 (sic), which is no longer necessary.
705 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
706 be the same on different OS.
708 PH/09 Moved a debug statement in filter processing to avoid a race problem when
711 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
712 whether --show-vars was specified or not
714 JJ/02 exipick: Added support for new ACL variable spool format introduced
717 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
718 syntactically invalid From: or Reply-to: line, and a filter used this to
719 generate an autoreply, and therefore failed to obtain an address for the
720 autoreply, Exim could try to deliver to a non-existent relative file
721 name, causing unrelated and misleading errors. What now happens is that
722 it logs this as a hard delivery error, but does not attempt to create a
725 PH/11 The exinext utility has a -C option for testing purposes, but although
726 the given file was scanned by exinext itself; it wasn't being passed on
727 when Exim was called.
729 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
730 an end-of-file indication when reading a command response.
732 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
733 compiled. In many other places in Exim, IPv6 addresses are always
734 recognized, so I have changed this. It also means that IPv4 domain
735 literals of the form [IPV4:n.n.n.n] are now always recognized.
737 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
738 used if the router is not running as root, for example, when verifying at
739 ACL time, or when using -bh. The debugging output from this situation was
740 non-existent - all you got was a failure to exec. I have made two
743 (a) Failures to set uid/gid, the current directory, or a process leader
744 in a subprocess such as that created by queryprogram now generate
745 suitable debugging ouput when -d is set.
747 (b) The queryprogram router detects when it is not running as root,
748 outputs suitable debugging information if -d is set, and then runs
749 the subprocess without attempting to change uid/gid.
751 PH/15 Minor change to Makefile for building test_host (undocumented testing
754 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
755 additional section of a DNS packet that returns MX or SRV records.
756 Instead, it always explicitly searches for A/AAAA records. This avoids
757 major problems that occur when a DNS server includes only records of one
758 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
759 fixed another bug: if SRV records were looked up and the corresponding
760 address records were *not* found in the additional section, the port
761 values from the SRV records were lost.
763 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
764 using the correct key (the original address) when searching the retry
765 rules in order to find which one to use for generating the retry hint.
767 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
768 adding the default one. Similarly, if it contains a Reply-To: header, the
769 errors_reply_to option, if set, is not used.
771 PH/19 When calculating a retry time, Exim used to measure the "time since
772 failure" by looking at the "first failed" field in the retry record. Now
773 it does not use this if it is later than than the arrival time of the
774 message. Instead it uses the arrival time. This makes for better
775 behaviour in cases where some deliveries succeed, thus re-setting the
776 "first failed" field. An example is a quota failure for a huge message
777 when small messages continue to be delivered. Without this change, the
778 "time since failure" will always be short, possible causing more frequent
779 delivery attempts for the huge message than are intended.
780 [Note: This change was subsequently modified - see PH/04 for 4.62.]
782 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
783 $1, $2, $3) because the numerical variables can be reset during some
784 expansion items (e.g. "match"), thereby losing the authentication data.
786 PH/21 Make -bV show the size of off_t variables so that the test suite can
787 decide whether to run tests for quotas > 2G.
789 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
790 mailbox_size, and mailbox_filecount in the appendfile transport. If a
791 filecount value is greater than 2G or if a quota value is greater than 2G
792 on a system where the size of off_t is not greater than 4, a panic error
795 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
796 never match. The debug and -bh output now contains an explicit error
797 message indicating a malformed IPv4 address or mask.
799 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
800 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
803 PH/25 Do not write to syslog when running in the test harness. The only
804 occasion when this arises is a failure to open the main or panic logs
805 (for which there is an explicit test).
807 PH/26 Added the /no_tell option to "control=freeze".
809 PH/27 If a host name lookup failed very early in a connection, for example, if
810 the IP address matched host_lookup and the reverse lookup yielded a name
811 that did not have a forward lookup, an error message of the form "no IP
812 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
813 could be logged. Now it outputs the IP address instead of "NULL".
815 PH/28 An enabling patch from MH: add new function child_open_exim2() which
816 allows the sender and the authenticated sender to be set when
817 submitting a message from within Exim. Since child_open_exim() is
818 documented for local_scan(), the new function should be too.
820 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
821 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
822 results in an empty string is now treated as unset.
824 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
826 PH/31 Added sender_verify_fail logging option.
828 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
829 needed by RFC 822 but not by RFC 2822 was commented out. I have now
830 tidied the source and removed it altogether.
832 PH/33 When a queue run was abandoned because the load average was too high, a
833 log line was always written; now it is written only if the queue_run log
834 selector is set. In addition, the log line for abandonment now contains
835 information about the queue run such as the pid. This is always present
836 in "start" and "stop" lines but was omitted from the "abandon" line.
838 PH/34 Omit spaces between a header name and the colon in the error message that
839 is given when verify = headers_syntax fails (if there are lots of them,
840 the message gets confusing).
842 PH/35 Change the default for dns_check_names_pattern to allow slashes within
843 names, as there are now some PTR records that contain slashes. This check
844 is only to protect against broken name servers that fall over on strange
845 characters, so the fact that it applies to all lookups doesn't matter.
847 PH/36 Now that the new test suite is complete, we can remove some of the
848 special code in Exim that was needed for the old test suite. For example,
849 sorting DNS records because real resolvers return them in an arbitrary
850 order. The new test suite's fake resolver always returns records in the
853 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
854 bounces) except when queue_only is set, to avoid logging races between
855 the different processes.
857 PH/38 Panic-die if .include specifies a non-absolute path.
859 PH/39 A tweak to the "H" retry rule from its user.
861 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
862 a label. They prevented compilation on older perls.
864 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
865 a warning to be raised on newish perls.
867 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
868 on queue. Changes to match documented behaviour of showing count of
869 messages matching specified criteria.
871 PH/40 Changed the default ident timeout from 30s to 5s.
873 PH/41 Added support for the use of login_cap features, on those BSD systems
874 that have them, for controlling the resources used by pipe deliveries.
876 PH/42 The content-scanning code uses fopen() to create files in which to put
877 message data. Previously it was not paying any attention to the mode of
878 the files. Exim runs with umask(0) because the rest of the code creates
879 files with open(), and sets the required mode explicitly. Thus, these
880 files were ending up world-writeable. This was not a big issue, because,
881 being within the spool directory, they were not world-accessible. I have
882 created a function called modefopen, which takes an additional mode
883 argument. It sets umask(777), creates the file, chmods it to the required
884 mode, then resets the umask. All the relevant calls to fopen() in the
885 content scanning code have been changed to use this function.
887 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
888 to 24 hours. This avoids potential overflow problems when processing G
889 and H retry rules. I suspect nobody ever tinkers with this value.
891 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
893 PH/45 When the plaintext authenticator is running as a client, the server's
894 challenges are checked to ensure they are valid base64 strings. By
895 default, the authentication attempt is cancelled if an invalid string is
896 received. Setting client_ignore_invalid_base64 true ignores these errors.
897 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
898 they are received. Thus, the responses can be made to depend on the
899 challenges. If an invalid string is ignored, an empty string is placed in
902 PH/46 Messages that are created by the autoreply transport now contains a
903 References: header, in accordance with RFCs 2822 and 3834.
905 PH/47 Added authenticated_sender_force to the smtp transport.
907 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
909 PH/49 Installed latest patch from the Sieve maintainer.
911 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
912 was also set, the appendfile transport was unnecessarily scanning a
913 directory of message files (e.g. for maildir delivery) to find the count
914 of files (along with the size), even though it did not need this
915 information. It now does the scan only if it needs to find either the
916 size of the count of files.
918 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
920 PH/52 Two bugs concerned with error handling when the smtp transport is
923 (i) Exim was not creating retry information for temporary errors given
924 for individual recipients after the DATA command when the smtp transport
925 was used in LMTP mode. This meant that they could be retried too
926 frequently, and not timed out correctly.
928 (ii) Exim was setting the flag that allows error details to be returned
929 for LMTP errors on RCPT commands, but not for LMTP errors for individual
930 recipients that were returned after the DATA command.
932 PH/53 This is related to PH/52, but is more general: for any failing address,
933 when detailed error information was permitted to be returned to the
934 sender, but the error was temporary, then after the final timeout, only
935 "retry timeout exceeded" was returned. Now it returns the full error as
936 well as "retry timeout exceeded".
938 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
939 do this, and (what is worse) MTAs that accept it.
941 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
942 will now be deprecated.
944 PH/56 New os.c-cygwin from the Cygwin maintainer.
946 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
947 formats (previously only available in exim formats via -bpr, -bpru,
948 and -bpra. Now also available in native and exiqgrep formats)
950 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
951 with very large, slow to parse queues
953 JJ/08 exipick: added ! as generic prefix to negate any criteria format
955 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
957 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
958 responses to authentication challenges, though it was showing the
959 challenges; (ii) I've removed the CR characters from the debug output for
962 PH/58 Allow for the insertion of a newline as well as a space when a string
963 is turned into more than one encoded-word during RFC 2047 encoding. The
964 Sieve code now uses this.
966 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
967 data_4xx, lost_connection, tls_required.
969 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
970 message was being sent as an SMTP response.
972 PH/61 Add -l and -k options to exicyclog.
974 PH/62 When verifying, if an address was redirected to one new address, so that
975 verification continued, and the new address failed or deferred after
976 having set something in $address_data, the value of $address_data was not
977 passed back to the ACL. This was different to the case when no
978 redirection occurred. The value is now passed back in both cases.
980 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
981 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
982 use login_cap.h, so on its own it isn't the distinguishing feature. The
983 new name refers directly to the setclassresources() function.
985 PH/65 Added configuration files for NetBSD3.
987 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
989 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
990 is preferred over IPv4.
992 PH/68 The bounce_return_message and bounce_return_body options were not being
993 honoured for bounces generated during the reception of non-SMTP messages.
994 In particular, this applied to messages rejected by the ACL. This bug has
995 been fixed. However, if bounce_return_message is true and bounce_return_
996 body is false, the headers that are returned for a non-SMTP message
997 include only those that have been read before the error was detected.
998 (In the case of an ACL rejection, they have all been read.)
1000 PH/69 The HTML version of the specification is now built in a directory called
1001 spec_html instead of spec.html, because the latter looks like a path with
1002 a MIME-type, and this confuses some software.
1004 PH/70 Catch two compiler warnings in sieve.c.
1006 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1007 function verify_get_ident() calls ip_connect() to connect a socket, but
1008 if the "connect()" function timed out, ip_connect() used to close the
1009 socket. However, verify_get_ident() also closes the socket later, and in
1010 between Exim writes to the log, which may get opened at this point. When
1011 the socket was closed in ip_connect(), the log could get the same file
1012 descriptor number as the socket. This naturally causes chaos. The fix is
1013 not to close the socket in ip_connect(); the socket should be closed by
1014 the function that creates it. There was only one place in the code where
1015 this was missing, in the iplookup router, which I don't think anybody now
1016 uses, but I've fixed it anyway.
1018 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1019 well as to direct DNS lookups. Otherwise the handling of names in host
1020 lists is inconsistent and therefore confusing.
1026 PH/01 Two changes to the default runtime configuration:
1028 (1) Move the checks for relay_from_hosts and authenticated clients from
1029 after to before the (commented out) DNS black list checks.
1031 (2) Add control=submission to the relay_from_hosts and authenticated
1032 clients checks, on the grounds that messages accepted by these
1033 statements are most likely to be submissions.
1035 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1037 (1) Generate an error if the third argument for the ${prvs expansion is
1040 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1043 (3) Reset the variables that are obtained from the first argument of
1044 ${prvscheck and used in the second argument before leaving the code,
1045 because their memory is reclaimed, so using them afterwards may do
1048 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1049 one (it's much easier than Tom thought :-).
1051 (5) Because of (4), we can now allow for the use of $prvscheck_result
1052 inside the third argument.
1054 PH/03 For some reason, the default setting of PATH when running a command from
1055 a pipe transport was just "/usr/bin". I have changed it to
1058 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1059 anything to be listed in the output from -bV.
1061 PH/05 When a filter generated an autoreply, the entire To: header line was
1062 quoted in the delivery log line, like this:
1064 => >A.N.Other <ano@some.domain> <original@ddress> ...
1066 This has been changed so that it extracts the operative address. There
1067 may be more than one such address. If so, they are comma-separated, like
1070 => >ano@some.domain,ona@other.domain <original@ddress> ...
1072 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1073 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1074 not being looked up in the rDNS to get a host name, Exim was showing the
1075 IP address twice in Received: lines, even though the IP addresses were
1076 identical. For example:
1078 Received: from [1.2.3.4] (helo=[1.2.3.4])
1080 However, if the real host name was known, it was omitting the HELO data
1081 if it matched the actual IP address. This has been tidied up so that it
1082 doesn't show the same IP address twice.
1084 PH/07 When both +timestamp and +memory debugging was on, the value given by
1085 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1086 called by the debug printing, thereby overwriting the timestamp buffer.
1087 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1089 PH/08 When the original message was included in an autoreply transport, it
1090 always said "this is a copy of the message, including all the headers",
1091 even if body_only or headers_only was set. It now gives an appropriate
1094 PH/09 Applied a patch from the Sieve maintainer which:
1096 o fixes some comments
1097 o adds the (disabled) notify extension core
1098 o adds some debug output for the result of if/elsif tests
1099 o points to the current vacation draft in the documentation
1100 and documents the missing references header update
1104 o fixes a bug in processing the envelope test (when testing
1105 multiple envelope elements, the last element determinted the
1108 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1109 Electronic Mail") by including:
1111 Auto-submitted: auto-generated
1113 in the messages that it generates (bounce messages and others, such as
1114 warnings). In the case of bounce messages for non-SMTP mesages, there was
1115 also a typo: it was using "Auto_submitted" (underscore instead of
1116 hyphen). Since every message generated by Exim is necessarily in response
1117 to another message, thes have all been changed to:
1119 Auto-Submitted: auto-replied
1121 in accordance with these statements in the RFC:
1123 The auto-replied keyword:
1125 - SHOULD be used on messages sent in direct response to another
1126 message by an automatic process,
1128 - MUST NOT be used on manually-generated messages,
1130 - MAY be used on Delivery Status Notifications (DSNs) and Message
1131 Disposition Notifications (MDNs),
1133 - MUST NOT be used on messages generated by automatic or periodic
1134 processes, except for messages which are automatic responses to
1137 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1138 to the default Received: header definition.
1140 PH/12 Added log selector acl_warn_skipped (default on).
1142 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1143 variables because (a) they are in the wrong storage pool and (b) even if
1144 they were copied, it wouldn't work properly because of the caching.
1146 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1147 checking when decoding. Apparently there are clients that generate
1148 overlong encoded strings. Why am I not surprised?
1150 PH/15 If the first argument of "${if match_address" was not empty, but did not
1151 contain an "@" character, Exim crashed. Now it writes a panic log message
1152 and treats the condition as false.
1154 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1156 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1157 extension "envelope-auth". The code is finished and in agreement with
1158 other implementations, but there is no documentation so far and in fact,
1159 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1160 not changing the active code.
1162 Print executed "if" and "elsif" statements when debugging is used. This
1163 helps a great deal to understand what a filter does.
1165 Document more things not specified clearly in RFC3028. I had all this
1166 sorted out, when out of a sudden new issues came to my mind. Oops."
1168 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1171 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1172 canonical form (as documented). However, after a host name lookup from
1173 the IP address, check_host() was doing a simple string comparison with
1174 addresses acquired from the DNS when checking that the found name did
1175 have the original IP as one of its addresses. Since any found IPv6
1176 addresses are likely to be in abbreviated form, the comparison could
1177 fail. Luckily, there already exists a function for doing the comparison
1178 by converting both addresses to binary, so now that is used instead of
1179 the text comparison.
1181 PH/20 There was another similar case to PH/19, when a complete host name was
1182 given in a host list; looking up its IP address could give an abbreviated
1183 form, whereas the current host's name might or might not be abbreviated.
1184 The same fix has been applied.
1190 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1191 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1194 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1195 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1197 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1199 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1200 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1201 Also, for "Auto-Submitted", treat anything other than "no" as
1202 non-personal, in accordance with RFC 3834. (Previously it treated
1203 anything starting "auto-" as non-personal.)
1205 TF/01 The control=submission/name=... option had a problem with syntax
1206 errors if the name included a slash character. The /name= option
1207 now slurps the rest of the string, so it can include any characters
1208 but it must come last in the list of options (after /sender_retain
1211 PH/05 Some modifications to the interface to the fake nameserver for the new
1219 TK/01 Added the "success_on_redirect" address verification option. See
1220 NewStuff for rationale and an example.
1222 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1224 PH/02 Patch to exigrep to allow it to work on syslog lines.
1226 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1227 fread() to skip over the body file's header line, because in Cygwin the
1228 header line is locked and is inaccessible.
1230 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1231 co-exist for some time) to make it clear that it is the Exim ID that is
1232 referenced, not the Message-ID: header line.
1234 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1235 string_format() function, because snprintf() does not exist on all
1238 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1239 ${stat: expansion item.
1241 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1242 protocol synchronization error", to keep the pedants happy.
1244 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1245 well as for IRIX systems, when gcc is being used. See the host.c source
1248 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1250 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1253 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1254 transport and to the smtp transport in LMTP mode.
1256 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1258 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1259 run a filter in a subprocess. This could lead to confusion in subsequent
1260 lookups in the parent process. There should also be a search_tidyup() at
1261 the end of the subprocess.
1263 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1264 only if the host matched helo_try_verify_hosts, which caused the
1265 verification to occur when the EHLO/HELO command was issued. The ACL just
1266 tested the remembered result. Now, if a previous verification attempt has
1267 not happened, "verify = helo" does it there and then.
1269 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1271 TK/03 Fix log output including CR from clamd.
1273 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1274 exist provoked a memory error which could cause a segfault.
1276 PH/15 Installed PCRE 6.2
1278 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1280 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1281 of the problem. Specifically, suggested +O2 rather than +O1 for the
1284 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1286 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1287 record, the port was not correctly logged when the outgoing_port log
1288 selector was set (it logged the transort's default port).
1290 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1291 fallback_hosts, and "hosts" in the smtp transport.
1293 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1294 host errors such as "Connection refused".
1296 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1297 authentication with radiusclient 0.4.9:
1299 - Error returned from rc_read_config was caught wrongly
1300 - Username/password not passed on to radius server due to wrong length.
1302 The presumption is that some radiusclient API changes for 4.51/PH/17
1303 were not taken care of correctly. The code is still untested by me (my
1304 Linux distribution still has 0.3.2 of radiusclient), but it was
1305 contributed by a Radius user.
1307 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1308 expanding the "port" option of the smtp transport.
1310 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1311 while reading a MIME header. Thanks to Tom Hughes for a patch.
1313 PH/24 Include config.h inside local_scan.h so that configuration settings are
1316 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1317 that in an ACL for RCPT (for example), you can examine exactly what was
1320 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1321 commands, but it was not correctly comparing the address with the actual
1322 client host address. Thus, it would show the EHLO address in Received:
1323 header lines when this was not necessary.
1325 PH/27 Added the % operator to ${eval:}.
1327 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1328 it should be ignoring failures (because with -C, for example, it has lost
1329 privilege). It wasn't ignoring creation failures other than "already
1332 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1335 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1336 because an input line was too long, either on its own, or by virtue of
1337 too many continuations, the temporary file was not being removed, and the
1338 return code was incorrect.
1340 PH/31 Missing "BOOL" in function definition in filtertest.c.
1342 PH/32 Applied Sieve patches from the maintainer.
1344 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1346 PH/33 Added "verify = not_blind".
1348 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1349 Local/Makefile (with some defaults set). These are used in built scripts
1350 such as exicyclog, but they have never been used in the exim_install
1351 script (though there are many overriding facilities there). I have
1352 arranged that the exim_install script now takes note of these two
1355 PH/35 Installed configuration files for Dragonfly.
1357 PH/36 When a locally submitted message by a trusted user did not contain a
1358 From: header, and the sender address was obtained from -f or from an SMTP
1359 MAIL command, and the trusted user did not use -F to supply a sender
1360 name, $originator_name was incorrectly used when constructing a From:
1361 header. Furthermore, $originator_name was used for submission mode
1362 messages from external hosts without From: headers in a similar way,
1363 which is clearly wrong.
1365 PH/37 Added control=suppress_local_fixups.
1367 PH/38 When log_selector = +received_sender was set, and the addition of the
1368 sender made the log line's construction buffer exactly full, or one byte
1369 less than full, an overflow happened when the terminating "\n" was
1372 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1373 when the result of a list match is failure because a DNS lookup failed.
1375 PH/40 RM_COMMAND is now used in the building process.
1377 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1378 the "build-* directories that it finds.
1380 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1381 address in a domain literal was a prefix of an interface address.
1383 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1384 when verifying a sender address, unless rewrite_headers is false.
1386 PH/44 Wrote a long comment about why errors_to addresses are verified as
1387 recipients, not senders.
1389 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1390 the ratelimit ACL was added.
1392 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1394 PH/46 Added extra information about PostgreSQL errors to the error string.
1396 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1397 avoiding the need to install special zones in a real server. This is
1398 backwards compatible; if it can't find the fake resolver, it drops back.
1399 Thus, both old and new test suites can be run.
1401 TF/02 Added util/ratelimit.pl
1403 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1404 clock is set back in time.
1406 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1407 Candler <B.Candler@pobox.com>.
1409 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1410 OK for addresses that are the result of redirections.
1412 PH/48 A number of further additions for the benefit of the new test suite,
1413 including a fake gethostbyname() that interfaces to the fake DNS resolver
1416 TF/06 The fix for widen_domains has also been applied to qualify_single and
1417 search_parents which are the other dnslookup options that can cause
1420 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1423 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1425 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1426 service" if it encountered an MX record with an empty target hostname.
1427 The message is now "an MX or SRV record indicated no SMTP service".
1429 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1430 if the DNS of the sending site is misconfigured. This is quite a
1431 common situation. This change restores the behaviour of treating a
1432 helo verification defer as a failure.
1434 PH/51 If self=fail was set on a router, the bounce message did not include the
1435 actual error message.
1441 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1443 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1444 command itself ended in error, the underlying message about the transport
1445 filter timeout was being overwritten with the pipe command error. Now the
1446 underlying error message should be appended to the second error message.
1448 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1450 PH/02 Reduce the amount of output that "make" produces by default. Full output
1451 can still be requested.
1453 PH/03 The warning log line about a condition test deferring for a "warn" verb
1454 was being output only once per connection, rather than after each
1455 occurrence (because it was using the same function as for successful
1456 "warn" verbs). This seems wrong, so I have changed it.
1458 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1459 it should not have, which might have caused a crash in the right
1460 circumstances, but probably never did.
1462 PH/04 Installed a modified version of Tony Finch's patch to make submission
1463 mode fix the return path as well as the Sender: header line, and to
1464 add a /name= option so that you can make the user's friendly name appear
1467 TF/03 Added the control = fakedefer ACL modifier.
1469 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1470 Mark Lowes for thorough testing.
1472 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1474 TK/03 Merged latest SRS patch from Miles Wilton.
1476 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1477 with the definition in sysexits.h (which is #included earlier).
1478 Fortunately, Exim does not actually use EX_OK. The code used to try to
1479 preserve the sysexits.h value, by assumimg that macro definitions were
1480 scanned for macro replacements. I have been disabused of this notion,
1481 so now the code just undefines EX_OK before #including unistd.h.
1483 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1484 in the smtp transport. When a block could not be written in a single
1485 write() function, the timeout was being re-applied to each part-write.
1486 This seems wrong - if the receiver was accepting one byte at a time it
1487 would take for ever. The timeout is now adjusted when this happens. It
1488 doesn't have to be particularly precise.
1490 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1491 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1493 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1494 without a domain if the check to <postmaster@domain> fails.
1496 SC/01 Eximstats: added -xls and the ability to specify output files
1497 (patch written by Frank Heydlauf).
1499 SC/02 Eximstats: use FileHandles for outputing results.
1501 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1503 SC/04 Eximstats: fixed display of large numbers with -nvr option
1505 SC/05 Eximstats: fixed merging of reports with empty tables.
1507 SC/06 Eximstats: added the -include_original_destination flag
1509 SC/07 Eximstats: removed tabs and trailing whitespace.
1511 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1513 TK/06 MBOX spool code: Add real "From " MBOX separator line
1514 so the .eml file is really in mbox format (even though
1515 most programs do not really care). Patch from Alex Miller.
1517 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1518 The latter is generated from $received_to and is only set if the
1519 message has one envelope recipient. SA can use these headers,
1520 obviously out-of-the-box. Patch from Alex Miller.
1522 PH/08 The ${def test on a variable was returning false if the variable's
1523 value was "0", contrary to what the specification has always said!
1524 The result should be true unless the variable is empty.
1526 PH/09 The syntax error of a character other than { following "${if
1527 def:variable_name" (after optional whitespace) was not being diagnosed.
1528 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1529 accidental colon was present, for example, could give incorrect results.
1531 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1532 result is used (not including appendfile, where other changes are about
1535 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1536 This involved changing a lot of size variables from int to off_t. It
1537 should work with maildirs and everything.
1539 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1540 spamd dying while we are connected to it.
1542 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1545 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1546 function for BDB 4.3.
1548 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1550 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1551 the sources for PCRE have been re-arranged and more files are now
1554 PH/15 The code I had for printing potentially long long variables in PH/11
1555 above was not the best (it lost precision). The length of off_t variables
1556 is now inspected at build time, and an appropriate printing format (%ld
1557 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1558 to be "long long int" or "long int". This is needed for the internal
1559 formatting function string_vformat().
1561 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1562 the configuration file to be ":syslog", then the script "guesses" where
1563 the logs files are, rather than using the compiled in default. In our
1564 case the guess is not the same as the compiled default, so the script
1565 suddenly stopped working when I started to use syslog. The patch checks
1566 to see if log_file_path is "". If so, it attempts to read it from exim
1567 with no configuration file to get the compiled in version, before it
1568 falls back to the previous guessing code."
1570 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1571 implementing BATV in an Exim configuration. See NewStuff for the gory
1574 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1575 Makefile that are specific to HP-UX.
1577 PH/18 If the "use_postmaster" option was set for a recipient callout together
1578 with the "random" option, the postmaster address was used as the MAIL
1579 FROM address for the random test, but not for the subsequent recipient
1580 test. It is now used for both.
1582 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1583 patch removes a few documentation additions to RFC 3028, because the
1584 latest draft now contains them. It adds the new en;ascii-case comparator
1585 and a new error check for 8bit text in MIME parts. Comparator and
1586 require names are now matched exactly. I enabled the subaddress
1587 extension, but it is not well tested yet (read: it works for me)."
1589 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1590 rework some of the code of TK/09 above to avoid the hardwired use of
1591 "%lld" and "long long". Replaced the call to snprintf() with a call to
1594 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1595 records point to non-existent hosts", "retry timeout exceeded", and
1596 "retry time not reached for any host after a long failure period".
1598 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1599 experimental DomainKeys support:
1601 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1602 (2) On an error such as an illegally used "control", the wrong name for
1603 the control was given.
1605 These problems did NOT occur unless DomainKeys support was compiled.
1607 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1609 PH/24 Added ${if match_ip condition.
1611 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1612 databases so that it will be absolutely obvious if a crash occurs in the
1613 DB library. This is a regular occurrence (often caused by mis-matched
1616 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1617 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1618 picked up on a user's system that detects such things. There doesn't seem
1619 to be a gcc warning option for this - only an attribute that has to be
1620 put on the function's prototype. It seems that in Fedora Core 4 they have
1621 set this on a number of new functions. No doubt there will be more in due
1624 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1625 specify a transport. However, if an address that was verified by such a
1626 router was the subject of a callout, Exim crashed because it tried to
1627 read the rcpt_include_affixes from the non-existent transport. Now it
1628 just assumes that the setting of that option is false. This bug was
1629 introduced by 4.51/PH/31.
1631 PH/28 Changed -d+all to exclude +memory, because that information is very
1632 rarely of interest, but it makes the output a lot bigger. People tend to
1633 do -d+all out of habit.
1635 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1636 code in os-type was giving problems when libc.so lives in lib64, like on
1639 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1640 aren't the modern standard, and it seems that some systems' include files
1641 don't always have them. Exim was already checking for some of the newer
1642 ones like T_AAAA, and defining it itself. I've added checks for all the
1643 record types that Exim uses.
1645 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1646 not automatically generating a new one, as it is supposed to. This
1647 prevented TLS from working. If the file did exist, but contained invalid
1648 data, a new version was generated, as expected. It was only the case of a
1649 non-existent file that was broken.
1651 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1652 with a change in libdomainkeys > 0.64.
1654 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1655 from DNS. If the selector record carries the flag, it now has
1656 precedence over the domain-wide flag.
1658 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1660 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1661 the use of a port name that isn't defined in /etc/services) occurred, the
1662 message was deferred as in a normal delivery, and thus remained on the
1663 spool, instead of being failed because of the mua_wrapper setting. This
1664 is now fixed, and I tidied up some of the mua_wrapper messages at the
1667 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1668 the messages in a hash of arrays rather than using individual hashes.
1669 This is a bit cleaner and results in dramatic memory savings, albeit
1670 at a slight CPU cost.
1672 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1673 as requested by Marc Sherman.
1675 SC/10 Eximstats: added histograms for user specified patterns as requested
1678 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1680 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1681 fopen() in the content-scanning modules that did not already have it.
1687 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1688 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1690 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1692 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1694 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1695 to test Sieve filters that use "vacation".
1697 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1698 that changes the way the GnuTLS parameters are stored in the cache file.
1699 The new format can be generated externally. For backward compatibility,
1700 if the data in the cache doesn't make sense, Exim assumes it has read an
1701 old-format file, and it generates new data and writes a new file. This
1702 means that you can't go back to an older release without removing the
1705 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1706 work if there are any delivery delays because "one_time" forces the
1707 parent to be marked "delivered", so its unseen clone is never tried
1708 again. For this reason, Exim now forbids the simultaneous setting of
1711 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1712 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1713 entry if you want to know the details. The fix, however, neglected to
1714 consider the case when local delivery batching is involved. The test for
1715 "previously delivered" was not happening when checking to see if an
1716 address could be batched with a previous (undelivered) one; under
1717 certain circumstances this could lead to multiple deliveries to the same
1720 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1721 in its include files, and this causes problems building Exim.
1723 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1724 header_syntax) but Exim was just ignoring anything given after a slash.
1725 In particular, this caused confusion with an attempt to use "verify =
1726 reverse_host_lookup/defer_ok". An error is now given when options are
1727 supplied for verify items that do not have them. (Maybe reverse_host_
1728 lookup should have a defer_ok option, but that's a different point.)
1730 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1731 defined by RFC 821) to 2048, because there were problems with some AUTH
1732 commands, and RFC 1869 says the size should be increased for extended
1733 SMTP commands that take arguments.
1735 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1738 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1739 "unknown" error; now it says that the functionality isn't in the binary.
1741 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1742 an address' error message when a string expansion fails (syntax or
1743 whatever). Otherwise the password may appear in the log. Following change
1744 PH/42 below, there is no longer a chance of it appearing in a bounce
1747 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1749 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1750 of its addresses was ever tried. (Bugzilla bug #2.)
1752 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1753 the result incorrectly in the debug output. (It correctly added a newline
1754 to what was transported.)
1756 TF/01 Added $received_time.
1758 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1759 commented out examples of how to interface to a virus scanner and to
1760 SpamAssassin. Also added commented examples of av_scanner and
1761 spamd_address settings.
1763 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1764 and controls are allowed in which ACLs. There were a couple of minor
1765 errors. Some of the entries in the conditions table (which is a table of
1766 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1767 negation of where the condition IS allowed.
1769 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1771 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1772 header file does not have a version number, so I've had to invent a new
1773 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1774 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1775 radiusclient), but it was contributed by a Radius user.
1777 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1778 files or directories, for OpenSSL.
1780 PH/19 When an Exim process that is running as root has to create an Exim log
1781 file, it does so in a subprocess that runs as exim:exim so as to get the
1782 ownership right at creation (otherwise, other Exim processes might see
1783 the file with the wrong ownership). There was no test for failure of this
1784 fork() call, which would lead to the process getting stuck as it waited
1785 for a non-existent subprocess. Forks do occasionally fail when resources
1786 run out. I reviewed all the other calls to fork(); they all seem to check
1789 PH/20 When checking for unexpected SMTP input at connect time (before writing
1790 the banner), Exim was not dealing correctly with a non-positive return
1791 from the read() function. If the client had disconnected by this time,
1792 the result was a log entry for a synchronization error with an empty
1793 string after "input=" when read() returned zero. If read() returned -1
1794 (an event I could not check), uninitialized data bytes were printed.
1795 There were reports of junk text (parts of files, etc) appearing after
1798 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1800 PH/22 Added support for macro redefinition, and (re)definition in between
1801 driver and ACL definitions.
1803 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1804 forgetting to use the resulting value; it was using the unexpanded value.
1806 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1807 hadn't been configured. The fix is from Juergen Kreileder, who
1808 understands it better than I do:
1810 "Here's what I see happening with three configured cyrus_sasl
1811 authenticators configured (plain, login, cram-md5):
1813 On startup auth_cyrus_sasl_init() gets called for each of these.
1814 This means three calls to sasl_listmech() without a specified mech_list.
1815 => SASL tests which mechs of all available mechs actually work
1816 => three warnings about OTP not working
1817 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1819 With the patch, sasl_listmech() also gets called three times. But now
1820 SASL's mech_list option is set to the server_mech specified in the the
1821 authenticator. Or in other words, the answer from sasl_listmech()
1822 gets limited to just the mech you're testing for (which is different
1824 => the return list contains just 'plain' or 'login', 'cram-md5' or
1825 nothing depending on the value of ob->server_mech.
1827 I've just tested the patch: Authentication still works fine,
1828 unavailable mechs specified in the exim configuration are still
1829 caught, and the auth.log warnings about OTP are gone."
1831 PH/25 When debugging is enabled, the contents of the command line are added
1832 to the debugging output, even when log_selector=+arguments is not
1835 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1836 answer is "GNU", and only if the return is "GNU/something" is the answer
1839 PH/27 $acl_verify_message is now set immediately after the failure of a
1840 verification in an ACL, and so is available in subsequent modifiers. In
1841 particular, the message can be preserved by coding like this:
1843 warn !verify = sender
1844 set acl_m0 = $acl_verify_message
1846 Previously, $acl_verify_message was set only while expanding "message"
1847 and "log_message" when a very denied access.
1849 PH/28 Modified OS/os.c-Linux with
1851 -#ifndef OS_LOAD_AVERAGE
1852 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1854 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1855 nomenclature these days.)
1857 PH/29 Installed patch from the Sieve maintainer that adds the options
1858 sieve_useraddress and sieve_subaddress to the redirect router.
1860 PH/30 In these circumstances:
1861 . Two addresses routed to the same list of hosts;
1862 . First host does not offer TLS;
1863 . First host accepts first address;
1864 . First host gives temporary error to second address;
1865 . Second host offers TLS and a TLS session is established;
1866 . Second host accepts second address.
1867 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1868 and peerdn, if requested) that were in fact used only for the second
1871 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1872 attention to any local part prefix or suffix that was matched by the
1873 router that accepted the address. It now behaves in the same way as it
1874 does for delivery: the affixes are removed from the local part unless
1875 rcpt_include_affixes is set on the transport.
1877 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1878 timeout during the DATA phase of an incoming message.
1880 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1881 applied a patch sanctioned by the Sieve maintainer.
1883 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1884 the uid or gid is negative. A case of a negative gid caused this to be
1885 noticed. The fix allows for either to be negative.
1887 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1888 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1891 PH/36 The change PH/12 above was broken. Fixed it.
1893 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1894 the grounds that routing the same address twice would always produce the
1895 same answer. This might have been true once, but it is certainly no
1896 longer true now. Routing a child address may depend on the previous
1897 routing that produced that child. Some complicated redirection strategies
1898 went wrong when messages had multiple recipients, and made Exim's
1899 behaviour dependent on the order in which the addresses were given.
1901 I have moved the duplicate checking until after the routing is complete.
1902 Exim scans the addresses that are assigned to local and remote
1903 transports, and removes any duplicates. This means that more work will be
1904 done, as duplicates will always all be routed, but duplicates are
1905 presumably rare, so I don't expect this is of any significance.
1907 For deliveries to pipes, files, and autoreplies, the duplicate checking
1908 still happens during the routing process, since they are not going to be
1911 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1912 It corrects a timeout issue with spamd. This is Ian's comment: "The
1913 background is that sometimes spamd either never reads data from a
1914 connection it has accepted, or it never writes response data. The exiscan
1915 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1916 blindly assumes that writes won't block so it may never time out."
1918 PH/39 Allow G after quota size as well as K and M.
1920 PH/40 The value set for $authenticated_id in an authenticator may not contain
1921 binary zeroes or newlines because the value is written to log lines and
1922 to spool files. There was no check on this. Now the value is run through
1923 the string_printing() function so that such characters are converted to
1924 printable escape sequences.
1926 PH/41 $message_linecount is a new variable that contains the total number of
1927 lines in the message. Compare $body_linecount, which is the count for the
1930 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1931 bounce and delay warning messages, except in certain special cases, which
1934 (a) An SMTP error message from a remote host;
1935 (b) A message specified in a :fail: redirection;
1936 (c) A message specified in a "fail" command in a system filter;
1937 (d) A message specified in a FAIL return from the queryprogram router;
1938 (e) A message specified by the cannot_route_message router option.
1940 In these cases only, Exim does include the error details in bounce and
1941 warning messages. There are also a few cases where bland messages such
1942 as "unrouteable address" or "local delivery error" are given.
1944 PH/43 $value is now also set for the "else" part of a ${run expansion.
1946 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1947 being worked on, but at least Exim now implements the latest version to
1950 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1951 process to complete was treated as a delivery failure, a timeout while
1952 writing the message to the pipe was logged, but erroneously treated as a
1953 successful delivery. Such timeouts include transport filter timeouts. For
1954 consistency with the overall process timeout, these timeouts are now
1955 treated as errors, giving rise to delivery failures by default. However,
1956 there is now a new Boolean option for the pipe transport called
1957 timeout_defer, which, if set TRUE, converts the failures into defers for
1958 both kinds of timeout. A transport filter timeout is now identified in
1961 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1962 systems where "make" and "gmake" are different, calling "gmake" at top
1963 level broke things. I've arranged for the value of $(MAKE) to be passed
1964 from the Makefile to this script so that it can call the same version of
1968 A note about Exim versions 4.44 and 4.50
1969 ----------------------------------------
1971 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1972 changes of various kinds. As a consequence, a big documentation update was
1973 needed. This delayed the release for rather longer than seemed good, especially
1974 in the light of a couple of (minor) security issues. Therefore, the changes
1975 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1976 release. So 4.44 and 4.50 are in effect two different branches that both start
1979 I have left the 4.50 change log unchanged; it contains all the changes since
1980 4.43. The change log for 4.44 is below; many of its items are identical to
1981 those for 4.50. This seems to be the most sensible way to preserve the
1982 historical information.
1988 1. Minor wording change to the doc/README.SIEVE file.
1990 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1991 computation of the current number of files was incorrect.
1993 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1994 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1997 4. Give more explanation in the error message when the command for a transport
1998 filter fails to execute.
2000 5. There are several places where Exim runs a non-Exim command in a
2001 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2002 was being done only for the command run by the queryprogram router. It is
2003 now done for all such subprocesses. The other cases are: ${run, transport
2004 filters, and the commands run by the lmtp and pipe transports.
2006 6. Added CONFIGURE_GROUP build-time option.
2008 7. Some older OS have a limit of 256 on the maximum number of file
2009 descriptors. Exim was using setrlimit() to set 1000 as a large value
2010 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2011 systems. I've change it so that if it can't get 1000, it tries for 256.
2013 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2014 was an oversight, and furthermore, ever since the addition of extra
2015 controls (e.g. 4.43/32), the checks on when to allow different forms of
2016 "control" were broken. There should now be diagnostics for all cases when a
2017 control that does not make sense is encountered.
2019 9. Added the /retain_sender option to "control=submission".
2021 10. $recipients is now available in the predata ACL (oversight).
2023 11. Tidy the search cache before the fork to do a delivery from a message
2024 received from the command line. Otherwise the child will trigger a lookup
2025 failure and thereby defer the delivery if it tries to use (for example) a
2026 cached ldap connection that the parent has called unbind on.
2028 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2029 of $address_data from the recipient verification was clobbered by the
2030 sender verification.
2032 13. The value of address_data from a sender verification is now available in
2033 $sender_address_data in subsequent conditions in the ACL statement.
2035 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2037 15. Added a new option "connect=<time>" to callout options, to set a different
2040 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2041 was its contents. (It was OK if the option was not defined at all.)
2043 17. A "Completed" log line is now written for messages that are removed from
2044 the spool by the -Mrm option.
2046 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2047 information about exactly what failed.
2049 19. Added -dd to debug only the daemon process.
2051 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2052 handles timeouts, both on the server side and network timeouts. Renamed the
2053 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2055 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2056 It is now set to "smtps".
2058 22. $host_address is now set to the target address during the checking of
2059 ignore_target_hosts.
2061 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2062 being passed; this would have caused $sender_host_name to have been used if
2063 matching the list had actually called for a host name (not very likely,
2064 since this list is usually IP addresses). A host name is now passed as
2067 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2068 code that specifies a non-restarting handler (typically sigaction() in
2069 modern systems) in an attempt to fix a rare and obscure crash bug.
2071 25. Narrowed the window for a race in the daemon that could cause it to ignore
2072 SIGCHLD signals. This is not a major problem, because they are used only to
2073 wake it up if nothing else does.
2075 26. A malformed maildirsize file could cause Exim to calculate negative values
2076 for the mailbox size or file count. Odd effects could occur as a result.
2077 The maildirsize information is now recalculated if the size or filecount
2080 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2081 support for a long time. Removed HAVE_SYS_VFS_H.
2083 28. Installed the latest version of exipick from John Jetmore.
2085 29. In an address list, if the pattern was not a regular expression, an empty
2086 subject address (from a bounce message) matched only if the pattern was an
2087 empty string. Non-empty patterns were not even tested. This was the wrong
2088 because it is perfectly reasonable to use an empty address as part of a
2089 database query. An empty address is now tested by patterns that are
2090 lookups. However, all the other forms of pattern expect the subject to
2091 contain a local part and a domain, and therefore, for them, an empty
2092 address still always fails if the pattern is not itself empty.
2094 30. Exim went into a mad DNS loop when attempting to do a callout where the
2095 host was specified on an smtp transport, and looking it up yielded more
2096 than one IP address.
2098 31. Re-factored the code for checking spool and log partition space into a
2099 function that finds that data and another that does the check. The former
2100 is then used to implement four new variables: $spool_space, $log_space,
2101 $spool_inodes, and $log_inodes.
2103 32. The RFC2047 encoding function was originally intended for short strings
2104 such as real names; it was not keeping to the 75-character limit for
2105 encoded words that the RFC imposes. It now respects the limit, and
2106 generates multiple encoded words if necessary. To be on the safe side, I
2107 have increased the buffer size for the ${rfc2047: expansion operator from
2110 33. It is now permitted to omit both strings after an "if" condition; if the
2111 condition is true, the result is "true". As before, when the second string
2112 is omitted, a false condition yields an empty string. This makes it less
2113 cumbersome to write custom ACL and router conditions.
2115 34. Failure to deliver a bounce message always caused it to be frozen, even if
2116 there was an errors_to setting on the router. The errors_to setting is now
2119 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2120 canonical form (fully expanded) before being placed in
2121 $sender_host_address.
2123 36. The table in the code that translates DNS record types into text (T_A to
2124 "A" for instance) was missing entries for NS and CNAME. It is just possible
2125 that this could have caused confusion if both these types were looked up
2126 for the same domain, because the text type is used as part of Exim's
2127 per-process caching. But the chance of anyone hitting this buglet seems
2130 37. The dnsdb lookup has been extended in a number of ways.
2132 (1) There is a new type, "zns", which walks up the domain tree until it
2133 finds some nameserver records. It should be used with care.
2135 (2) There is a new type, "mxh", which is like "mx" except that it returns
2136 just the host names, not the priorities.
2138 (3) It is now possible to give a list of domains (or IP addresses) to be
2139 looked up. The behaviour when one of the lookups defers can be
2140 controlled by a keyword.
2142 (4) It is now possible to specify the separator character for use when
2143 multiple records are returned.
2145 38. The dnslists ACL condition has been extended: it is now possible to supply
2146 a list of IP addresses and/or domains to be looked up in a particular DNS
2149 39. Added log_selector=+queue_time_overall.
2151 40. When running the queue in the test harness, wait just a tad after forking a
2152 delivery process, to get repeatability of debugging output.
2154 41. Include certificate and key file names in error message when GnuTLS fails
2155 to set them up, because the GnuTLS error message doesn't include the name
2156 of the failing file when there is a problem reading it.
2158 42. Allow both -bf and -bF in the same test run.
2160 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2162 44. The "Exiscan patch" is now merged into the mainline Exim source.
2164 45. Sometimes the final signoff response after QUIT could fail to get
2165 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2166 < 0 before doing a fflush(). This bug looks as though it goes back to the
2167 introduction of TLS in release 3.20, but "sometimes" must have been rare
2168 because the tests only now provoked it.
2170 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2171 (this can affect the format of dates).
2173 47. exim_tidydb, when checking for the continued existence of a message for
2174 which it has found a message-specific retry record, was not finding
2175 messages that were in split spool directories. Consequently, it was
2176 deleting retry records that should have stayed in existence.
2178 48. Steve fixed some bugs in eximstats.
2180 49. The SPA authentication driver was not abandoning authentication and moving
2181 on to the next authenticator when an expansion was forced to fail,
2182 contradicting the general specification for all authenticators. Instead it
2183 was generating a temporary error. It now behaves as specified.
2185 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2186 (the order specifies the preference for clients). The order is now AES256,
2187 AES128, 3DES, ARCFOUR128.
2189 51. Small patch to Sieve code - explicitly set From: when generating an
2192 52. Exim crashed if a remote delivery caused a very long error message to be
2193 recorded - for instance if somebody sent an entire SpamAssassin report back
2194 as a large number of 550 error lines. This bug was coincidentally fixed by
2195 increasing the size of one of Exim's internal buffers (big_buffer) that
2196 happened as part of the Exiscan merge. However, to be on the safe side, I
2197 have made the code more robust (and fixed the comments that describe what
2200 53. Now that there can be additional text after "Completed" in log lines (if
2201 the queue_time_overall log selector is set), a one-byte patch to exigrep
2202 was needed to allow it to recognize "Completed" as not the last thing in
2205 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2206 patch that reportedly fixes this has been added. I am not expert enough to
2207 create a test for it. This is what the patch creator wrote:
2209 "I found a little strange behaviour of ldap code when working with
2210 Windows 2003 AD Domain, where users was placed in more than one
2211 Organization Units. When I tried to give exim partial DN, the exit code
2212 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2213 But simultaneously result of request was absolutely normal ldap result,
2214 so I produce this patch..."
2216 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2217 REFERENCE, so I have modified the code to exclude the patch when that macro
2220 55. Some experimental protocols are using DNS PTR records for new purposes. The
2221 keys for these records are domain names, not reversed IP addresses. The
2222 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2223 leaves it alone. Component reversal etc. now happens only for IP addresses.
2226 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2228 57. Double the size of the debug message buffer (to 2048) so that more of very
2229 long debug lines gets shown.
2231 58. The exicyclog utility now does better if the number of log files to keep
2232 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2234 59. Two changes related to the smtp_active_hostname option:
2236 (1) $smtp_active_hostname is now available as a variable.
2237 (2) The default for smtp_banner uses $smtp_active_hostname instead
2238 of $primary_hostname.
2240 60. The host_aton() function is supposed to be passed a string that is known
2241 to be a valid IP address. However, in the case of IPv6 addresses, it was
2242 not checking this. This is a hostage to fortune. Exim now panics and dies
2243 if the condition is not met. A case was found where this could be provoked
2244 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2245 components; fortuitously, this particular loophole had already been fixed
2246 by change 4.50/55 above.
2248 If there are any other similar loopholes, the new check in host_aton()
2249 itself should stop them being exploited. The report I received stated that
2250 data on the command line could provoke the exploit when Exim was running as
2251 exim, but did not say which command line option was involved. All I could
2252 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2253 running as the user.
2256 61. There was a buffer overflow vulnerability in the SPA authentication code
2257 (which came originally from the Samba project). I have added a test to the
2258 spa_base64_to_bits() function which I hope fixes it.
2261 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2262 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2264 63. The daemon start-up calls getloadavg() while still root for those OS that
2265 need the first call to be done as root, but it missed one case: when
2266 deliver_queue_load_max is set with deliver_drop_privilege. This is
2267 necessary for the benefit of the queue runner, because there is no re-exec
2268 when deliver_drop_privilege is set.
2270 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2271 This has been fixed.
2273 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2274 was in use, was not putting the data itself into the right store pool;
2275 consequently, it could be overwritten for a subsequent message in the same
2276 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2279 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2281 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2282 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2283 the function were treating the return as a boolean value, which happened to
2284 work because 0=false and not-0=true, but is not correct code.
2286 68. The host_aton() function was not handling scoped IPv6 addresses (those
2287 with, for example, "%eth0" on the end) correctly.
2289 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2290 negated items (that is, ~something) in unsigned ints. Some compilers
2291 apparently mutter when there is no cast.
2293 70. If an address verification called from an ACL failed, and did not produce a
2294 user-specific message (i.e. there was only a "system" message), nothing was
2295 put in $acl_verify_message. In this situation, it now puts the system
2298 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2299 session; change 4.31/43 added the unwanted input to the log line - except
2300 that it did not do this in the start of session case. It now does.
2302 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2303 This is wrong and can cause the other end to generate a synchronization
2304 error if it is another Exim or anything else that does the synchronization
2305 check. A QUIT command is no longer sent after a timeout.
2307 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2308 during host lookups.
2310 74. The defer_ok option of callout verification was not working if it was used
2311 when verifying addresses in header lines, that is, for this case:
2313 verify = header_sender/callout=defer_ok
2315 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2316 those file descriptors could be used for SMTP connections. If anything
2317 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2318 could be sent to the SMTP client, causing chaos. The daemon now opens
2319 stdin, stdout, and stderr to /dev/null when it puts itself into the
2322 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2323 log by default. The user can override this with suitable Perl magic.
2325 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2326 the log message when discard triggers, was not working for the DATA ACL or
2327 for the non-SMTP ACL.
2329 78. Error message wording change in sieve.c.
2331 79. If smtp_accept_max_per_host was set, the number of connections could be
2332 restricted to fewer than expected, because the daemon was trying to set up
2333 a new connection before checking whether the processes handling previous
2334 connections had finished. The check for completed processes is now done
2335 earlier. On busy systems, this bug wouldn't be noticed because something
2336 else would have woken the daemon, and it would have reaped the completed
2339 80. If a message was submitted locally by a user whose login name contained one
2340 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2341 It caused a spool format error. I have fixed the spool reading code. A
2342 related problem was that the "from" clause in the Received: line became
2343 illegal because of the space(s). It is now covered by ${quote_local_part.
2345 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2348 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2350 83. Patch from Sieve maintainer to fix unterminated string problem in
2351 "vacation" handling.
2353 84. Some minor changes to the Linux configuration files to help with other
2354 OS variants using glibc.
2356 85. One more patch for Sieve to update vacation handling to latest spec.
2359 ----------------------------------------------------
2360 See the note above about the 4.44 and 4.50 releases.
2361 ----------------------------------------------------
2367 1. Change 4.43/35 introduced a bug that caused file counts to be
2368 incorrectly computed when quota_filecount was set in an appendfile
2371 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2372 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2375 3. Give more explanation in the error message when the command for a transport
2376 filter fails to execute.
2378 4. There are several places where Exim runs a non-Exim command in a
2379 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2380 was being done only for the command run by the queryprogram router. It is
2381 now done for all such subprocesses. The other cases are: ${run, transport
2382 filters, and the commands run by the lmtp and pipe transports.
2384 5. Some older OS have a limit of 256 on the maximum number of file
2385 descriptors. Exim was using setrlimit() to set 1000 as a large value
2386 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2387 systems. I've change it so that if it can't get 1000, it tries for 256.
2389 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2390 was an oversight, and furthermore, ever since the addition of extra
2391 controls (e.g. 4.43/32), the checks on when to allow different forms of
2392 "control" were broken. There should now be diagnostics for all cases when a
2393 control that does not make sense is encountered.
2395 7. $recipients is now available in the predata ACL (oversight).
2397 8. Tidy the search cache before the fork to do a delivery from a message
2398 received from the command line. Otherwise the child will trigger a lookup
2399 failure and thereby defer the delivery if it tries to use (for example) a
2400 cached ldap connection that the parent has called unbind on.
2402 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2403 of $address_data from the recipient verification was clobbered by the
2404 sender verification.
2406 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2407 was its contents. (It was OK if the option was not defined at all.)
2409 11. A "Completed" log line is now written for messages that are removed from
2410 the spool by the -Mrm option.
2412 12. $host_address is now set to the target address during the checking of
2413 ignore_target_hosts.
2415 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2416 being passed; this would have caused $sender_host_name to have been used if
2417 matching the list had actually called for a host name (not very likely,
2418 since this list is usually IP addresses). A host name is now passed as
2421 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2422 code that specifies a non-restarting handler (typically sigaction() in
2423 modern systems) in an attempt to fix a rare and obscure crash bug.
2425 15. Narrowed the window for a race in the daemon that could cause it to ignore
2426 SIGCHLD signals. This is not a major problem, because they are used only to
2427 wake it up if nothing else does.
2429 16. A malformed maildirsize file could cause Exim to calculate negative values
2430 for the mailbox size or file count. Odd effects could occur as a result.
2431 The maildirsize information is now recalculated if the size or filecount
2434 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2435 support for a long time. Removed HAVE_SYS_VFS_H.
2437 18. Updated exipick to current release from John Jetmore.
2439 19. Allow an empty sender to be matched against a lookup in an address list.
2440 Previously the only cases considered were a regular expression, or an
2443 20. Exim went into a mad DNS lookup loop when doing a callout where the
2444 host was specified on the transport, if the DNS lookup yielded more than
2447 21. The RFC2047 encoding function was originally intended for short strings
2448 such as real names; it was not keeping to the 75-character limit for
2449 encoded words that the RFC imposes. It now respects the limit, and
2450 generates multiple encoded words if necessary. To be on the safe side, I
2451 have increased the buffer size for the ${rfc2047: expansion operator from
2454 22. Failure to deliver a bounce message always caused it to be frozen, even if
2455 there was an errors_to setting on the router. The errors_to setting is now
2458 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2459 canonical form (fully expanded) before being placed in
2460 $sender_host_address.
2462 24. Updated eximstats to version 1.33
2464 25. Include certificate and key file names in error message when GnuTLS fails
2465 to set them up, because the GnuTLS error message doesn't include the name
2466 of the failing file when there is a problem reading it.
2468 26. Expand error message when OpenSSL has problems setting up cert/key files.
2471 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2472 (this can affect the format of dates).
2474 28. exim_tidydb, when checking for the continued existence of a message for
2475 which it has found a message-specific retry record, was not finding
2476 messages that were in split spool directories. Consequently, it was
2477 deleting retry records that should have stayed in existence.
2479 29. eximstats updated to version 1.35
2480 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2481 1.35 - bugfix such that pie charts by volume are generated correctly
2483 30. The SPA authentication driver was not abandoning authentication and moving
2484 on to the next authenticator when an expansion was forced to fail,
2485 contradicting the general specification for all authenticators. Instead it
2486 was generating a temporary error. It now behaves as specified.
2488 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2489 (the order specifies the preference for clients). The order is now AES256,
2490 AES128, 3DES, ARCFOUR128.
2492 31. Small patch to Sieve code - explicitly set From: when generating an
2495 32. Exim crashed if a remote delivery caused a very long error message to be
2496 recorded - for instance if somebody sent an entire SpamAssassin report back
2497 as a large number of 550 error lines. This bug was coincidentally fixed by
2498 increasing the size of one of Exim's internal buffers (big_buffer) that
2499 happened as part of the Exiscan merge. However, to be on the safe side, I
2500 have made the code more robust (and fixed the comments that describe what
2503 33. Some experimental protocols are using DNS PTR records for new purposes. The
2504 keys for these records are domain names, not reversed IP addresses. The
2505 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2506 leaves it alone. Component reversal etc. now happens only for IP addresses.
2509 34. The host_aton() function is supposed to be passed a string that is known
2510 to be a valid IP address. However, in the case of IPv6 addresses, it was
2511 not checking this. This is a hostage to fortune. Exim now panics and dies
2512 if the condition is not met. A case was found where this could be provoked
2513 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2514 components; fortuitously, this particular loophole had already been fixed
2515 by change 4.50/55 or 4.44/33 above.
2517 If there are any other similar loopholes, the new check in host_aton()
2518 itself should stop them being exploited. The report I received stated that
2519 data on the command line could provoke the exploit when Exim was running as
2520 exim, but did not say which command line option was involved. All I could
2521 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2522 running as the user.
2525 35. There was a buffer overflow vulnerability in the SPA authentication code
2526 (which came originally from the Samba project). I have added a test to the
2527 spa_base64_to_bits() function which I hope fixes it.
2530 36. The daemon start-up calls getloadavg() while still root for those OS that
2531 need the first call to be done as root, but it missed one case: when
2532 deliver_queue_load_max is set with deliver_drop_privilege. This is
2533 necessary for the benefit of the queue runner, because there is no re-exec
2534 when deliver_drop_privilege is set.
2536 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2537 was in use, was not putting the data itself into the right store pool;
2538 consequently, it could be overwritten for a subsequent message in the same
2539 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2542 38. Sometimes the final signoff response after QUIT could fail to get
2543 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2544 < 0 before doing a fflush(). This bug looks as though it goes back to the
2545 introduction of TLS in release 3.20, but "sometimes" must have been rare
2546 because the tests only now provoked it.
2552 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2553 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2554 Now it returns an integer. A number of calls were still expecting a T/F
2555 return. Fortuitously, in all cases, the tests worked in OK situations,
2556 which is the norm. However, things would have gone wrong on any write
2557 failures on the smtp file descriptor. This function is used when sending
2558 messages over SMTP and also when doing verify callouts.
2560 2. When Exim is called to do synchronous delivery of a locally submitted
2561 message (the -odf or -odi options), it no longer closes stderr before doing
2564 3. Implemented the mua_wrapper option.
2566 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2568 5. Implemented the functions header_remove(), header_testname(),
2569 header_add_at_position(), and receive_remove_recipient(), and exported them
2572 6. If an ACL "warn" statement specified the addition of headers, Exim already
2573 inserted X-ACL-Warn: at the start if there was no header name. However, it
2574 was not making this test for the second and subsequent header lines if
2575 there were newlines in the string. This meant that an invalid header could
2576 be inserted if Exim was badly configured.
2578 7. Allow an ACL "warn" statement to add header lines at the start or after all
2579 the Received: headers, as well as at the end.
2581 8. Added the rcpt_4xx retry error code.
2583 9. Added postmaster_mailfrom=xxx to callout verification option.
2585 10. Added mailfrom=xxxx to the callout verification option, for verify=
2588 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2589 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2591 12. Inserted some casts to stop certain compilers warning when using pointer
2592 differences as field lengths or precisions in printf-type calls (mostly
2593 affecting debugging statements).
2595 13. Added optional readline() support for -be (dynamically loaded).
2597 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2598 same clock tick as a message's arrival, so that its received time was the
2599 same as the "first fail" time on the retry record, and that message
2600 remained on the queue past the ultimate address timeout, every queue runner
2601 would try a delivery (because it was past the ultimate address timeout) but
2602 after another failure, the ultimate address timeout, which should have then
2603 bounced the address, did not kick in. This was a "< instead of <=" error;
2604 in most cases the first failure would have been in the next clock tick
2605 after the received time, and all would be well.
2607 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2608 being recognized when the domain list was tested by the match_domain
2609 condition in an expansion string.
2611 16. Added the ${str2b64: operator.
2613 17. Exim was always calling setrlimit() to set a large limit for the number of
2614 processes, without checking whether the existing limit was already
2615 adequate. (It did check for the limit on file descriptors.) Furthermore,
2616 errors from getrlimit() and setrlimit() were being ignored. Now they are
2617 logged to the main and panic logs, but Exim does carry on, to try to do its
2618 job under whatever limits there are.
2620 18. Imported PCRE 5.0.
2622 19. Trivial typo in log message " temporarily refused connection" (the leading
2625 20. If the log selector return_path_on_delivery was set and an address was
2626 redirected to /dev/null, the delivery process crashed because it assumed
2627 that a return path would always be set for a "successful" delivery. In this
2628 case, the whole delivery is bypassed as an optimization, and therefore no
2631 21. Internal re-arrangement: the function for sending a challenge and reading
2632 a response while authentication was assuming a zero-terminated challenge
2633 string. It's now changed to take a pointer and a length, to allow for
2634 binary data in such strings.
2636 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2638 23. Exim was not respecting finduser_retries when seeking the login of the
2639 uid under which it was called; it was always trying 10 times. (The default
2640 setting of finduser_retries is zero.) Also, it was sleeping after the final
2641 failure, which is pointless.
2643 24. Implemented tls_on_connect_ports.
2645 25. Implemented acl_smtp_predata.
2647 26. If the domain in control=submission is set empty, Exim assumes that the
2648 authenticated id is a complete email address when it generates From: or
2649 Sender: header lines.
2651 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2652 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2653 chown and chgrp in /bin and hostname in /usr/bin.
2655 28. Exim was keeping the "process log" file open after each use, just as it
2656 does for the main log. This opens the possibility of it remaining open for
2657 long periods when the USR1 signal hits a daemon. Occasional processlog
2658 errors were reported, that could have been caused by this. Anyway, it seems
2659 much more sensible not to leave this file open at all, so that is what now
2662 29. The long-running daemon process does not normally write to the log once it
2663 has entered its main loop, and it closes the log before doing so. This is
2664 so that log files can straightforwardly be renamed and moved. However,
2665 there are a couple of unusual error situations where the daemon does write
2666 log entries, and I had neglected to close the log afterwards.
2668 30. The text of an SMTP error response that was received during a remote
2669 delivery was being truncated at 512 bytes. This is too short for some of
2670 the long messages that one sometimes sees. I've increased the limit to
2673 31. It is now possible to make retry rules that apply only when a message has a
2674 specific sender, in particular, an empty sender.
2676 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2677 it possible to be selective about when SMTP synchronization is enforced.
2679 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2681 32. Implemented hosts_connection_nolog.
2683 33. Added an ACL for QUIT.
2685 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2688 35. Added mailbox_size and mailbox_filecount to appendfile.
2690 36. Added control = no_multiline_responses to ACLs.
2692 37. There was a bug in the logic of the code that waits for the clock to tick
2693 in the case where the clock went backwards by a substantial amount such
2694 that the microsecond fraction of "now" was more than the microsecond
2695 fraction of "then" (but the whole seconds number was less).
2697 38. Added support for the libradius Radius client library this is found on
2698 FreeBSD (previously only the radiusclient library was supported).
2704 1. When certain lookups returned multiple values in the form name=value, the
2705 quoting of the values was not always being done properly. Specifically:
2706 (a) If the value started with a double quote, but contained no whitespace,
2708 (b) If the value contained whitespace other than a space character (i.e.
2709 tabs or newlines or carriage returns) it was not quoted.
2710 This fix has been applied to the mysql and pgsql lookups by writing a
2711 separate quoting function and calling it from the lookup code. The fix
2712 should probably also be applied to nisplus, ibase and oracle lookups, but
2713 since I cannot test any of those, I have not disturbed their existing code.
2715 2. A hit in the callout cache for a specific address caused a log line with no
2716 reason for rejecting RCPT. Now it says "Previous (cached) callout
2717 verification failure".
2719 3. There was an off-by-one bug in the queryprogram router. An over-long
2720 return line was truncated at 256 instead of 255 characters, thereby
2721 overflowing its buffer with the terminating zero. As well as fixing this, I
2722 have increased the buffer size to 1024 (and made a note to document this).
2724 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2725 when Exim is waiting for an SMTP response from a remote server, Exim
2726 restarts its select() call on the socket, thereby resetting its timeout.
2727 This is not a problem when such interrupts are rare. Somebody set up a cron
2728 job to run exiwhat every 2 minutes, which is less than the normal select()
2729 timeout (5 or 10 minutes). This meant that the select() timeout never
2730 kicked in because it was always reset. I have fixed this by comparing the
2731 time when an interrupt arrives with the time at the start of the first call
2732 to select(). If more time than the timeout has elapsed, the interrupt is
2733 treated as a timeout.
2735 5. Some internal re-factoring in preparation for the addition of Sieve
2736 extensions (by MH). In particular, the "personal" test is moved to a
2737 separate function, and given an option for scanning Cc: and Bcc: (which is
2738 not set for Exim filters).
2740 6. When Exim created an email address using the login of the caller as the
2741 local part (e.g. when creating a From: or Sender: header line), it was not
2742 quoting the local part when it contained special characters such as @.
2744 7. Installed new OpenBSD configuration files.
2746 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2747 try to make them clearer.
2749 9. Callout options, other than the timeout value, were being ignored when
2750 verifying sender addresses in header lines. For example, when using
2752 verify = header_sender/callout=no_cache
2754 the cache was (incorrectly) being used.
2756 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2757 only the Cygwin environment.
2759 11. When return_path_on_delivery was set as a log selector, if different remote
2760 addresses in the same message used different return paths and parallel
2761 remote delivery occurred, the wrong values would sometimes be logged.
2762 (Whenever a remote delivery process finished, the return path value from
2763 the most recently started remote delivery process was logged.)
2765 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2766 lines when AUTH and/or TLS are in use. This is the "received protocol"
2767 field. Exim used to use "asmtp" for authenticated SMTP, without any
2768 indication (in the protocol name) for TLS use. Now it follows the RFC and
2769 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2770 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2771 names appear in log lines as well as in Received: header lines.
2773 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2774 extensions, and comparison tests, and to fix some bugs.
2776 14. Changes to the "personal" filter test:
2778 (1) The test was buggy in that it was just doing the equivalent of
2779 "contains" tests on header lines. For example, if a user's address was
2780 anne@some.where, the "personal" test would incorrectly be true for
2782 To: susanne@some.where
2784 This test is now done by extracting each address from the header in turn,
2785 and checking the entire address. Other tests that are part of "personal"
2786 are now done using regular expressions (for example, to check local parts
2787 of addresses in From: header lines).
2789 (2) The list of non-personal local parts in From: addresses has been
2790 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2791 taken from the Sieve specification recommendations.
2793 (3) If the message contains any header line starting with "List-" it is
2794 treated as non-personal.
2796 (4) The test for "circular" in the Subject: header line has been removed
2797 because it now seems ill-conceived.
2799 15. Minor typos in src/EDITME comments corrected.
2801 16. Installed latest exipick from John Jetmore.
2803 17. If headers_add on a router specified a text string that was too long for
2804 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2805 of string_sprintf() is now avoided.
2807 18. $message_body_size was not set (it was always zero) when running the DATA
2808 ACL and the local_scan() function.
2810 19. For the "mail" command in an Exim filter, no default was being set for
2811 the once_repeat time, causing a random time value to be used if "once" was
2812 specified. (If the value happened to be <= 0, no repeat happened.) The
2813 default is now 0s, meaning "never repeat". The "vacation" command was OK
2814 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2815 (I found it when inspecting the code).
2817 20. There is now an overall timeout for performing a callout verification. It
2818 defaults to 4 times the callout timeout, which applies to individual SMTP
2819 commands during the callout. The overall timeout applies when there is more
2820 than one host that can be tried. The timeout is checked before trying the
2821 next host. This prevents very long delays if there are a large number of
2822 hosts and all are timing out (e.g. when the network connections are timing
2823 out). The value of the overall timeout can be changed by specifying an
2824 additional sub-option for "callout", called "maxwait". For example:
2826 verify = sender/callout=5s,maxwait=20s
2828 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2829 to the end before writing, but this should make it even safer).
2831 22. Exim was forgetting that it had advertised PIPELINING for the second and
2832 subsequent messages on an SMTP connection. It was also not resetting its
2833 memory on STARTTLS and an internal HELO.
2835 23. When Exim logs an SMTP synchronization error within a session, it now
2836 records whether PIPELINING has been advertised or not.
2838 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2839 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2840 rather than long int.
2842 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2844 26. Added the never_mail option to autoreply.
2850 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2851 crash if the getsockname() call failed; this can happen if a connection is
2852 closed very soon after it is established. The problem was simply in the
2853 order in which certain operations were done, causing Exim to try to write
2854 to the SMTP stream before it had set up the file descriptor. The bug has
2855 been fixed by making things happen in the correct order.
2861 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2862 before the connection was closed, thus losing the rejection response.
2864 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2865 some early Solaris releases, but causes trouble in current releases where
2866 socklen_t is defined.
2868 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2871 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2874 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2875 that releasing the top of it at the end releases what was used for sub-
2876 expansions (unless the block got too big). However, discard this block if
2877 the first thing is a variable or header, so that we can use its block when
2878 it is dynamic (useful for very large $message_headers, for example).
2880 6. Lookups now cache *every* query, not just the most recent. A new, separate
2881 store pool is used for this. It can be recovered when all lookup caches are
2882 flushed. Lookups now release memory at the end of their result strings.
2883 This has involved some general refactoring of the lookup sources.
2885 7. Some code has been added to the store_xxx() functions to reduce the amount
2886 of flapping under certain conditions.
2888 8. log_incoming_interface used to affect only the <= reception log lines. Now
2889 it causes the local interface and port to be added to several more SMTP log
2890 lines, for example "SMTP connection from", and rejection lines.
2892 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2894 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2896 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2897 could be overwritten at the end of the current message (or the start of a
2898 new message if it was set in a HELO ACL). The value is now preserved for
2899 the duration of the SMTP connection.
2901 12. If a transport had a headers_rewrite setting, and a matching header line
2902 contained an unqualified address, that address was qualified, even if it
2903 did not match any rewriting rules. The underlying bug was that the values
2904 of the flags that permit the existence of unqualified sender and recipient
2905 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2906 non-local messages, and by -bnq for local messages) were not being
2907 preserved with the message after it was received.
2909 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2910 "next input=" as part of the text comprising the host identity instead of
2911 the correct text. The code was using the same buffer for two different
2912 strings. However, depending on which order the printing function evaluated
2913 its arguments, the bug did not always show up. Under Linux, for example, my
2914 test suite worked just fine.
2916 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2917 doesn't work with some older versions of Perl. It has been changed to "my",
2918 which in any case is probably the better facility to use.
2920 15. A really picky compiler found some instances of statements for creating
2921 error messages that either had too many or two few arguments for the format
2924 16. The size of the buffer for calls to the DNS resolver has been increased
2925 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2926 for addresses that have a lot of PTR records. This alleviates a problem; it
2927 does not fully solve it.
2929 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2930 buffer now truncates the list and logs the incident, which is the same
2931 action as happens when Exim is looking up a host name and its aliases.
2932 Previously in this situation something unpredictable would happen;
2933 sometimes it was "internal error: store_reset failed".
2935 18. If a server dropped the connection unexpectedly when an Exim client was
2936 using GnuTLS and trying to read a response, the client delivery process
2937 crashed while trying to generate an error log message.
2939 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2940 string, for example:
2942 warn message = H1: something\nH2: something
2944 the text was added as a single header line from Exim's point of view
2945 though it ended up OK in the delivered message. However, searching for the
2946 second and subsequent header lines using $h_h2: did not work. This has been
2947 fixed. Similarly, if a system filter added multiple headers in this way,
2948 the routers could not see them.
2950 20. Expanded the error message when iplsearch is called with an invalid key to
2951 suggest using net-iplsearch in a host list.
2953 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2954 ACLs are no longer actually imposed (and a message to that effect is
2957 22. If a "gecos" field in a passwd entry contained escaped characters, in
2958 particular, if it contained a \" sequence, Exim got it wrong when building
2959 a From: or a Sender: header from that name. A second bug also caused
2960 incorrect handling when an unquoted " was present following a character
2961 that needed quoting.
2963 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2964 was not being matched caselessly.
2966 24. Arranged for all hyphens in the exim.8 source to be escaped with
2969 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2970 properly. Recipient callout cache records were still being keyed to include
2971 the sender, even when use_sender was set false. This led to far more
2972 callouts that were necessary. The sender is no longer included in the key
2973 when use_sender is false.
2975 26. Added "control = submission" modifier to ACLs.
2977 27. Added the ${base62d: operator to decode base 62 numbers.
2979 28. dnsdb lookups can now access SRV records.
2981 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2982 the configuration file.
2984 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2985 (-v) mode. This makes the output for a verbose queue run more intelligible.
2987 31. Added a use_postmaster feature to recipient callouts.
2989 32. Added the $body_zerocount variable, containing the number of binary zero
2990 bytes in the message body.
2992 33. The time of last modification of the "new" subdirectory is now used as the
2993 "mailbox time last read" when there is a quota error for a maildir
2996 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2998 35. Added +ignore_unknown as a special item in host lists.
3000 36. Code for decoding IPv6 addresses in host lists is now included, even if
3001 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3002 address was recognized as an IP address, but was then not correctly decoded
3003 into binary, causing unexpected and incorrect effects when compared with
3010 1. Very minor rewording of debugging text in manualroute to say "list of
3011 hosts" instead of "hostlist".
3013 2. If verify=header_syntax was set, and a header line with an unqualified
3014 address (no domain) and a large number of spaces between the end of the
3015 name and the colon was received, the reception process suffered a buffer
3016 overflow, and (when I tested it) crashed. This was caused by some obsolete
3017 code that should have been removed. The fix is to remove it!
3019 3. When running in the test harness, delay a bit after writing a bounce
3020 message to get a bit more predictability in the log output.
3022 4. Added a call to search_tidyup() just before forking a reception process. In
3023 theory, someone could use a lookup in the expansion of smtp_accept_max_
3024 per_host which, without the tidyup, could leave open a database connection.
3026 5. Added the variables $recipient_data and $sender_data which get set from a
3027 lookup success in an ACL "recipients" or "senders" condition, or a router
3028 "senders" option, similar to $domain_data and $local_part_data.
3030 6. Moved the writing of debug_print from before to after the "senders" test
3033 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3034 problems for message scanning, either using a data ACL, or using
3035 local_scan() because the Received: header was not generated till after they
3036 were called (in order to set the time as the time of reception completion).
3037 I have revised the way this works. The header is now generated after the
3038 body is received, but before the ACL or local_scan() are called. After they
3039 are run, the timestamp in the header is updated.
3045 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3046 before starting a queue runner without re-exec. This happened only when
3047 deliver_drop_privilege was set or when the Exim user was set to root. The
3048 effect of the bug was that timeouts during subsequent deliveries caused
3049 crashes instead of being properly handled. The handler is now left at its
3050 default (and expected) setting.
3052 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3053 message, again when deliver_drop_privilege is set or Exim is run as root.
3054 The bug described in (1) was not present in this case, but the tidying up
3055 of the other signals was missing. I have made the two cases consistent.
3057 3. The ignore_target_hosts setting on a manualroute router was being ignored
3058 for hosts that were looked up using the /MX notation.
3060 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3063 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3064 operated on the sender address. After changing the $sender_address to <>
3065 for the sender address verify, Exim was re-instated it as the original
3066 (before rewriting) address, but remembering that it had rewritten it, so it
3067 wasn't rewriting it again. This bug also had the effect of breaking the
3068 sender address verification caching when the sender address was rewritten.
3070 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3071 This has been changed so that if the ip literal address matches
3072 ignore_target_hosts, the router declines.
3074 7. Added expansion conditions match_domain, match_address, and match_local_
3075 part (NOT match_host).
3077 8. The placeholder for the Received: header didn't have a length field set.
3079 9. Added code to Exim itself and to exim_lock to test for a specific race
3080 condition that could lead to file corruption when using MBX delivery. The
3081 issue is with the lockfile that is created in /tmp. If this file is removed
3082 after a process has opened it but before that process has acquired a lock,
3083 there is the potential for a second process to recreate the file and also
3084 acquire a lock. This could lead to two Exim processes writing to the file
3085 at the same time. The added code performs the same test as UW imapd; it
3086 checks after acquiring the lock that its file descriptor still refers to
3087 the same named file.
3089 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3090 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3093 11. Added the smtp_active_hostname option. If used, this will typically be made
3094 to depend on the incoming interface address. Because $interface_address is
3095 not set up until the daemon has forked a reception process, error responses
3096 that can happen earlier (such as "too many connections") no longer contain
3099 12. If an expansion in a condition on a "warn" statement fails because a lookup
3100 defers, the "warn" statement is abandoned, and the next ACL statement is
3101 processed. Previously this caused the whole ACL to be aborted.
3103 13. Added the iplsearch lookup type.
3105 14. Added ident_timeout as a log selector.
3107 15. Added tls_certificate_verified as a log selector.
3109 16. Added a global option tls_require_ciphers (compare the smtp transport
3110 option of the same name). This controls incoming TLS connections.
3112 17. I finally figured out how to make tls_require_ciphers do a similar thing
3113 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3114 before starting the TLS session.
3116 18. Tabs are now shown as \t in -bP output.
3118 19. If the log selector return_path_on_delivery was set, Exim crashed when
3119 bouncing a message because it had too many Received: header lines.
3121 20. If two routers both had headers_remove settings, and the first one included
3122 a superfluous trailing colon, the final name in the first list and the
3123 first name in the second list were incorrectly joined into one item (with a
3124 colon in the middle).
3130 1. Added -C and -D options to the exinext utility, mainly to make it easier
3131 to include in the automated testing, but these could be helpful when
3132 multiple configurations are in use.
3134 2. The exinext utility was not formatting the output nicely when there was
3135 an alternate port involved in the retry record key, nor when there was a
3136 message id as well (for retries that were specific to a specific message
3137 and a specific host). It was also confused by IPv6 addresses, because of
3138 the additional colons they contain. I have fixed the IPv4 problem, and
3139 patched it up to do a reasonable job for IPv6.
3141 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3142 delivery, the log line now contains "pipelined" if PIPELINING was used.
3144 4. An SMTP transport process used to panic and die if the bind() call to set
3145 an explicit outgoing interface failed. This has been changed; it is now
3146 treated in the same way as a connect() failure.
3148 5. A reference to $sender_host_name in the part of a conditional expansion
3149 that was being skipped was still causing a DNS lookup. This no longer
3152 6. The def: expansion condition was not recognizing references to header lines
3153 that used bh_ and bheader_.
3155 7. Added the _cache feature to named lists.
3157 8. The code for checking quota_filecount in the appendfile transport was
3158 allowing one more file than it should have been.
3160 9. For compatibility with Sendmail, the command line option
3168 and sets the incoming protocol and host name (for trusted callers). The
3169 host name and its colon can be omitted when only the protocol is to be set.
3170 Note the Exim already has two private options, -pd and -ps, that refer to
3171 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3172 "s", but I don't think that's a major issue.
3174 10. A number of refactoring changes to the code, none of which should affect
3177 (a) The number of logging options was getting close to filling up the
3178 32-bit word that was used as a bit map. I have split them into two classes:
3179 those that are passed in the argument to log_write(), and those that are
3180 only ever tested independently outside of that function. These are now in
3181 separate 32-bit words, so there is plenty of room for expansion again.
3182 There is no change in the user interface or the logging behaviour.
3184 (b) When building, for example, log lines, the code previously used a
3185 macro that called string_cat() twice, in order to add two strings. This is
3186 not really sufficiently general. Furthermore, there was one instance where
3187 it was actually wrong because one of the argument was used twice, and in
3188 one call a function was used. (As it happened, calling the function twice
3189 did not affect the overall behaviour.) The macro has been replaced by a
3190 function that can join an arbitrary number of extra strings onto a growing
3193 (c) The code for expansion conditions now uses a table and a binary chop
3194 instead of a serial search (which was left over from when there were very
3195 few conditions). Also, it now recognizes conditions like "pam" even when
3196 the relevant support is not compiled in: a suitably worded error message is
3197 given if an attempt is made to use such a condition.
3199 11. Added ${time_interval:xxxxx}.
3201 12. A bug was causing one of the ddress fields not to be passed back correctly
3202 from remote delivery subprocesses. The field in question was not being
3203 subsequently used, so this caused to problems in practice.
3205 13. Added new log selectors queue_time and deliver_time.
3207 14. Might have fixed a bug in maildirsizefile handling that threw up
3208 "unexpected character" debug warnings, and recalculated the data
3209 unnecessarily. In any case, I expanded the warning message to give more
3212 15. Added the message "Restricted characters in address" to the statements in
3213 the default ACL that block characters like @ and % in local parts.
3215 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3216 Three changes have been made:
3218 (a) There was a serious bug; a negative response to MAIL caused the whole
3219 recipient domain to be cached as invalid, thereby blocking all messages
3220 to all local parts at the same domain, from all senders. This bug has
3221 been fixed. The domain is no longer cached after a negative response to
3222 MAIL if the sender used is not empty.
3224 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3227 (c) A new callout option, "use_sender" has been added for people who want
3228 the modified behaviour.
3234 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3237 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3238 indeed breaks things for older releases.
3240 3. Added additional logging to the case where there is a problem reading data
3241 from a filter that is running in a subprocess using a pipe, in order to
3242 try to track down a specific problem.
3244 4. Testing facility fudge: when running in the test harness and attempting
3245 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3246 getting "No route to host". Convert this to a timeout.
3248 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3251 6. Some OS don't have socklen_t but use size_t instead. This affects the
3252 fifth argument of getsockopt() amongst other things. This is now
3253 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3254 can be set for individual OS. I have set it for SunOS5, OSF1, and
3255 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3256 some earlier ones do not.
3258 7. Change 4.30/15 was not doing the test caselessly.
3260 8. The standard form for an IPv6 address literal was being rejected by address
3261 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3262 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3263 this, as well as the form without the "IPv6" on the front (but only when
3264 address literals are enabled, of course).
3266 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3268 10. Exim crashed if a message with an empty sender address specified by -f
3269 encountered a router with an errors_to setting. This could be provoked only
3270 by a command such as
3274 where an empty string was supplied; "<>" did not hit this bug.
3276 11. Installed PCRE release 4.5.
3278 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3279 remained set. It is now erased.
3281 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3282 times from message ids (which are base 36 rather than the normal 62).
3284 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3285 were being counted as actual protocol errors, and logged if the log
3286 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3287 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3288 and DATA following a set of rejected RCPTs do not count as protocol errors.
3289 In other words, Exim assumes they were pipelined, though this may not
3290 actually be the case. Of course, in all cases the client gets an
3291 appropriate error code.
3293 15. If a lookup fails in an ACL condition, a message about the failure may
3294 be available; it is used if testing the ACL cannot continue, because most
3295 such messages specify what the cause of the deferral is. However, some
3296 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3297 that caused an old message to be retained and used if a later statement
3298 caused a defer, replacing the real cause of the deferral.
3300 16. If an IP address had so many PTR records that the DNS lookup buffer
3301 was not large enough to hold them, Exim could crash while trying to process
3302 the truncated data. It now detects and logs this case.
3304 17. Further to 4.21/58, another change has been made: if (and only if) the
3305 first line of a message (the first header line) ends with CRLF, a bare LF
3306 in a subsequent header line has a space inserted after it, so as not to
3307 terminate the header.
3309 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3310 unnecessarily, used atoi() instead of strtol(), and didn't check the
3311 termination when getting file sizes from file names by regex.
3313 19. Completely re-implemented the support for maildirsize files, in the light
3314 of a number of problems with the previous contributed implementation
3315 (4.30/29). In particular:
3317 . If the quota is zero, the maildirsize file is maintained, but no quota is
3320 . If the maildir directory does not exist, it is created before any attempt
3321 to write a maildirsize file.
3323 . The quota value in the file is just a cache; if the quota is changed in
3324 the transport, the new value overrides.
3326 . A regular expression is available for excluding directories from the
3329 20. The autoreply transport checks the characters in options that define the
3330 message's headers; it allows continued headers, but it was checking with
3331 isspace() after an embedded newline instead of explicitly looking for a
3334 21. If all the "regular" hosts to which an address was routed had passed their
3335 expiry times, and had not reached their retry times, the address was
3336 bounced, even if fallback hosts were defined. Now Exim should go on to try
3339 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3340 equivalent code in the SMTP transport. Some hosts send humungous responses
3341 to HELO/EHLO, more than 1024 it seems.
3343 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3344 gives compiler warnings in some environments. I've now done it "properly",
3347 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3348 (because of problems with the built-in one) was declared to return uschar *
3349 instead of char *, causing compiler failure.
3351 25. Fixed a file descriptor leak when processing alias/forward files.
3353 26. Fixed a minor format string issue in dbfn.c.
3355 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3357 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3358 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3361 29. When a sender address is verified, it is cached, to save repeating the test
3362 when there is more than one recipient in a message. However, when the
3363 verification involves a callout, it is possible for different callout
3364 options to be set for different recipients. It is too complicated to keep
3365 track of this in the cache, so now Exim always runs a verification when a
3366 callout is required, relying on the callout cache for the optimization.
3367 The overhead is duplication of the address routing, but this should not be
3370 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3371 to be verified with callout=postmaster, and the main callout worked but the
3372 postmaster check failed, the verification correctly failed. However, if a
3373 subsequent RCPT command asked for sender verification *without* the
3374 postmaster check, incorrect caching caused this verification also to fail,
3377 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3378 it was not caching the DNS options (qualify_single, search_parents) that
3379 were used when the lookup failed. A subsequent lookup with different
3380 options therefore always gave the same answer, though there were cases
3381 where it should not have. (Example: a "domains = !$mx_any" option on a
3382 dnslookup router: the "domains" option is always processed without any
3383 widening, but the router might have qualify_single set.) Now Exim uses the
3384 cached value only when the same options are set.
3386 32. Added John Jetmore's "exipick" utility to the distribution.
3388 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3389 than a timeout (e.g. a certificate is required, and is not provided), an
3390 Exim server now closes the connection immediately. Previously it waited for
3391 the client to close - but if the client is SSL, it seems that they each
3392 wait for each other, leading to a delay before one of them times out.
3394 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3395 maintained 0.8.x compatibility because I don't think many are using it, and
3396 it is clearly obsolete.
3398 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3401 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3402 client certificate was expired. A simple patch fixes this, though I don't
3403 understand the full logic of why the verify callback is called multiple
3406 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3407 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3408 which causes problems with some clients (such as the Certicom SSL Plus
3409 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3410 disables the coutermeasure allowing Eudora to connect."
3412 38. Exim was not checking that a write() to a log file succeeded. This could
3413 lead to Bad Things if a log got too big, in particular if it hit a file
3414 size limit. Exim now panics and dies if it cannot write to a log file, just
3415 as it does if it cannot open a log file.
3417 39. Modified OS/Makefile-Linux so that it now contains
3419 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3421 The two -D definitions ensure that Exim is compiled with large file
3422 support, which makes it possible to handle log files that are bigger than
3425 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3426 instance) a domain was checked against a named list that involved a lookup,
3427 causing $domain_data to be set, then another domain was checked against the
3428 same list, then the first domain was re-checked, the value of $domain_data
3429 after the final check could be wrong. In particular, if the second check
3430 failed, it could be set empty. This bug probably also applied to
3433 41. The strip_trailing_dot option was not being applied to the address given
3434 with the -f command-line option.
3436 42. The code for reading a message's header from the spool was incrementing
3437 $received_count, but never initializing it. This meant that the value was
3438 incorrect (doubled) while delivering a message in the same process in which
3439 it was received. In the most common configuration of Exim, this never
3440 happens - a fresh exec is done - but it can happen when
3441 deliver_drop_privilege is set.
3443 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3444 it now includes up to 150 characters of the unexpected data in the log
3447 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3448 and building data strings. The size of both of these buffers was 10 000
3449 bytes - far larger than anybody would *ever* want, thought I. Needless to
3450 say, somebody hit the limit. I have increased the maximum line length to
3451 20 000 and the maximum data length of concatenated lines to 100 000. I have
3452 also fixed two bugs, because there was no checking on these buffers. Tsk,
3453 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3454 buffer is too small.
3456 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3457 lsearch lookups. Now it does.
3459 46. When parsing a route_list item in a manualroute router, a fixed-length
3460 buffer was used for the list of hosts. I made this 1024 bytes long,
3461 thinking that nobody would ever have a list of hosts that long. Wrong.
3462 Somebody had a whole pile of complicated expansion conditions, and the
3463 string was silently truncated, leading to an expansion error. It turns out
3464 that it is easier to change to an unlimited length (owing to other changes
3465 that have happened since this code was originally written) than to build
3466 structure for giving a limitation error. The length of the item that
3467 expands into the list of hosts is now unlimited.
3469 47. The lsearch lookup could not handle data where the length of text line was
3470 more than 4095 characters. Such lines were truncated, leading to shortened
3471 data being returned. It should now handle lines of any length.
3473 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3474 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3477 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3478 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3479 no longer happens in comment lines. A list of replacements is now placed
3480 at the head of all of the source files, except those whose only change is
3481 to replace PERL_COMMAND in the very first #! line.
3483 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3484 messages on the queue, with a bottom-up merge sort, using code contributed
3485 by Michael Haardt. This should make operations like -bp somewhat faster on
3486 large queues. It won't affect queue runners, except when queue_run_in_order
3489 51. Installed eximstats 1.31 in the distribution.
3491 52. Added support for SRV lookups to the dnslookup router.
3493 53. If an ACL referred to $message_body or $message_body_end, the value was not
3494 reset for any messages that followed in the same SMTP session.
3496 54. The store-handling optimization for building very long strings was not
3497 differentiating between the different store pools. I don't think this
3498 actually made any difference in practice, but I've tidied it.
3500 55. While running the routers to verify a sender address, $sender_address
3501 was still set to the sender address. This is wrong, because when routing to
3502 send a bounce to the sender, it would be empty. Therefore, I have changed
3503 it so that, while verifying a sender address, $sender_address is set to <>.
3504 (There is no change to what happens when verifying a recipient address.)
3506 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3507 target A or AAAA records (if not already returned) without resetting the
3508 qualify_single or search_parents options of the DNS resolver. These are
3509 inappropriate in this case because the targets of MX and SRV records must
3510 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3511 target that, when qualified, matched something in the local domain. These
3512 two options are now turned off when doing these lookups.
3514 57. It seems that at least some releases of Reiserfs (which does not have the
3515 concept of a fixed number of inodes) returns zero and not -1 for the
3516 number of available inodes. This interacted badly with check_spool_inodes,
3517 which assumed that -1 was the "no such thing" setting. What I have done is
3518 to check that the total number of inodes is greater than zero before doing
3519 the test of how many are available.
3521 58. When a "warn" ACL statement has a log_message modifier, the message is
3522 remembered, and not repeated. This is to avoid a lot of repetition when a
3523 message has many recipients that cause the same warning to be written.
3524 Howewer, Exim was preserving the list of already written lines for an
3525 entire SMTP session, which doesn't seem right. The memory is now reset if a
3526 new message is started.
3528 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3529 debugging output unless log_rewrite was also set.
3531 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3532 of (int)(handle) when we know that handle contains (void *)(-1).
3534 61. The Exim daemon panic-logs an error return when it closes the incoming
3535 connection. However "connection reset by peer" seems to be common, and
3536 isn't really an error worthy of noting specially, so that particular error
3539 62. When Exim is trying to find all the local interfaces, it used to panic and
3540 die if the ioctl to get the interface flags failed. However, it seems that
3541 on at least one OS (Solaris 9) it is possible to have an interface that is
3542 included in the list of interfaces, but for which you get a failure error
3543 for this call. This happens when the interface is not "plumbed" into a
3544 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3545 failure of the "get flags" call assumes that the interface is down.
3547 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3548 makes life easier for people who are doing arithmetic on fields extracted
3549 from dates, where you often get leading zeros that should not be
3550 interpreted as octal.
3552 64. Added qualify_domain to the redirect router, to override the global
3555 65. If a pathologically long header line contained very many addresses (the
3556 report of this problem mentioned 10 000) and each of them was rewritten,
3557 Exim could use up a very large amount of memory. (It kept on making new
3558 copies of the header line as it rewrote, and never released the old ones.)
3559 At the expense of a bit more processing, the header rewriting function has
3560 been changed so that it no longer eats memory in this way.
3562 66. The generation of the Received: header has been moved from the time that a
3563 message starts to be received, to the time that it finishes. The timestamp
3564 in the Received: header should now be very close to that of the <= log
3565 line. There are two side-effects of this change:
3567 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3568 logged header lines no longer include the local Received: line, because
3569 it has not yet been created. The same applies to a copy of the message
3570 that is returned to a non-SMTP sender when a message is rejected.
3572 (b) When a filter file is tested using -bf, no additional Received: header
3573 is added to the test message. After some thought, I decided that this
3576 This change does not affect the value of $received_for. It is still set
3577 after address rewriting, but before local_scan() is called.
3579 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3581 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3582 gave an unhelpful panic error message, and a defer error. I have managed to
3583 change this behaviour so that it now rejects any supplied certificate,
3584 which seems right, as the list of acceptable certificates is empty.
3586 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3587 gave an unhelpful defer error. I have not managed to make this reject any
3588 supplied certificates, but the error message it gives is "no certificate
3589 supplied", which is not helpful.
3591 70. exigrep's output now also includes lines that are not associated with any
3592 message, but which match the given pattern. Implemented by a patch from
3593 Martin Sluka, which also tidied up the Perl a bit.
3595 71. Recipient callout verification, like sender verification, was using <> in
3596 the MAIL FROM command. This isn't really the right thing, since the actual
3597 sender may affect whether the remote host accepts the recipient or not. I
3598 have changed it to use the actual sender in the callout; this means that
3599 the cache record is now keyed on a recipient/sender pair, not just the
3600 recipient address. There doesn't seem to be a real danger of callout loops,
3601 since a callout by the remote host to check the sender would use <>.
3602 [SEE ABOVE: changed after hitting problems.]
3604 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3605 temporary errors. However, in the case of such a code being given after
3606 the end of a data transmission (i.e. after ".") Exim was failing to write
3607 a retry record for the message. (Yes, there was some broken host that was
3608 actually sending 8xx at this point.)
3610 73. An unknown lookup type in a host list could cause Exim to panic-die when
3611 the list was checked. (An example that provoked this was putting <; in the
3612 middle of a list instead of at the start.) If this happened during a DATA
3613 ACL check, a -D file could be left lying around. This kind of configuration
3614 error no longer causes Exim to die; instead it causes a defer errror. The
3615 incident is still logged to the main and panic logs.
3617 74. Buglet left over from Exim 3 conversion. The message "too many messages
3618 in one connection" was written to the rejectlog but not the mainlog, except
3619 when address rewriting (yes!) was being logged.
3621 75. Added write_rejectlog option.
3623 76. When a system filter was run not as root (that is, when system_filter_user
3624 was set), the values of the $n variables were not being returned to the
3625 main process; thus, they were not subsequently available in the $sn
3628 77. Added +return_path_on_delivery log selector.
3630 78. A connection timeout was being treated differently from recipients deferred
3631 when testing hosts_max_try with a message that was older than the host's
3632 retry timeout. (The host should not be counted, thus allowing all hosts to
3633 be tried at least once before bouncing.) This may have been the cause of an
3634 occasionally reported bug whereby a message would remain on the queue
3635 longer than the retry timeout, but would be bounced if a delivery was
3636 forced. I say "may" because I never totally pinned down the problem;
3637 setting up timeout/retry tests is difficult. See also the next item.
3639 79. The ultimate address timeout was not being applied to errors that involved
3640 a combination of host plus message (for example, a timeout on a MAIL
3641 command). When an address resolved to a number of possible hosts, and they
3642 were not all tried for each delivery (e.g. because of hosts_max_try), a
3643 message could remain on the queue longer than the retry timeout.
3645 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3648 81. Fixed an obscure SMTP outgoing bug which required at least the following
3649 conditions: (a) there was another message waiting for the same server;
3650 (b) the server returned 5xx to all RCPT commands in the first message so
3651 that the message was not completed; (c) the server dropped the connection
3652 or gave a negative response to the RSET that Exim sends to abort the
3653 transaction. The observed case was a dropped connection after DATA that had
3654 been sent in pipelining mode. That is, the server had advertised PIPELINING
3655 but was not implementing it correctly. The effect of the bug was incorrect
3656 behaviour, such as trying another host, and this could lead to a crash.
3662 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3663 and daemon.c were passed as pointers to ints; they should have been
3664 pointers to socklen_t variables (which are typically unsigned ints).
3666 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3669 3. Fixed a really odd bug that affected only the testing scheme; patching a
3670 certain fixed string in the binary changed the value of another string that
3671 happened to be identical to the end of the original first string.
3673 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3674 name", it returns that address as the IP address. On some operating
3675 systems (e.g. Solaris), it also passes back the IP address string as the
3676 "host name". However, on others (e.g. Linux), it passes back an empty
3677 string. Exim wasn't checking for this, and was changing the host name to an
3678 empty string, assuming it had been canonicized.
3680 5. Although rare, it is permitted to have more than one PTR record for a given
3681 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3682 all the names associated with an address, because they do in Solaris.
3683 However, it seems that they do not in Linux for data that comes from the
3684 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3685 I found this out when I moved to a new Linux workstation and tried to run
3686 the Exim test suite.
3688 To get round this problem I have changed the code so that it now does its
3689 own call to the DNS to look up PTR records when searching for a host name.
3690 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3691 addresses that are only in /etc/hosts are still found.
3693 This behaviour is, however, controlled by an option called host_lookup_
3694 order, which defaults to "bydns:byaddr". If people want to use the other
3695 order, or indeed, just use one or the other means of lookup, they can
3696 specify it in this variable.
3698 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3699 some operating systems, this comes back from gethostbyaddr() as an empty
3700 string, and this is what Exim used to test for. However, it seems that in
3701 other systems, "." is yielded. Exim now tests for this case too.
3703 7. The values of check_spool_space and check_log_space are now held internally
3704 as a number of kilobytes instead of an absolute number of bytes. If a
3705 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3706 kilobyte. This means that much larger values can be stored.
3708 8. Exim monitor: an attempt to get the action menu when not actually pointing
3709 at a message produces an empty menu entitled "No message selected". This
3710 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3711 no entries in it ("Shell widget menu has zero width and/or height"). So I
3712 have added a single, blank menu entry in this case.
3714 9. Added ${quote_local_part.
3716 10. MIME decoding is now applied to the contents of Subject: header lines when
3719 11. Now that a reference to $sender_host_address automatically causes a reverse
3720 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3721 host lookup before query-style lookups in lists that might use this
3722 variable. This has therefore been abolished, and the "net-" prefix is no
3723 longer necessary for query-style lookups.
3725 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3726 have been a typo for LFLAGS, so it has been changed.
3728 13. The install script calls Exim with "-C /dev/null" in order to find the
3729 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3730 to be output. Howeve, since Exim outputs its version number before the
3731 error, it didn't break the script. It just looked ugly. I fixed this by
3732 always allowing "-C /dev/null" if the caller is root.
3734 14. Ignore overlarge ACL variable number when reading spool file - insurance
3735 against a later release with more variables having written the file.
3737 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3738 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3739 this, as well as the form without the "IPv6" on the front.
3741 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3742 OS/Makefile-Darwin file.
3744 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3745 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3747 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3748 variable. However, some OS use ints for inodes. I've added cast to long int
3749 to get rid of the compiler warning.
3751 19. I had forgotten to lock out "/../" in configuration file names when
3752 ALT_CONFIG_PREFIX was set.
3754 20. Routers used for verification do not need to specify transports. However,
3755 if such a router generated a host list, and callout was configured, Exim
3756 crashed, because it could not find a port number from the (non-existent)
3757 transport. It now assumes port 25 in this circumstance.
3759 21. Added the -t option to exigrep.
3761 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3762 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3763 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3764 them is set, LOOKUP_LSEARCH is forced.
3766 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3768 24. Added sender and host information to the "rejected by local_scan()" log
3769 line; previously there was no indication of these.
3771 25. Added .include_if_exists.
3773 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3774 Linux. It turns out that not all file systems support this. Apparently some
3775 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3776 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3777 not supported on the file descriptor, is now ignored when Exim is trying to
3778 sync a directory. This applies only to Linux.
3780 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3782 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3783 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3786 29. Added support for maildirsize files from supplied patch (modified a bit).
3788 30. The use of :fail: followed by an empty string could lead Exim to respond to
3789 sender verification failures with (e.g.):
3791 550 Verification failed for <xxx>
3792 550 Sender verify failed
3794 where the first response line was missing the '-' that indicates it is not
3795 the final line of the response.
3797 31. The loop for finding the name of the user that called Exim had a hardwired
3798 limit of 10; it now uses the value of finduser_retries, which is used for
3799 all other user lookups.
3801 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3804 33. Exim was neglecting to zero errno before one call of strtol() when
3805 expanding a string and expecting an integer value. On some systems this
3806 resulted in spurious "integer overflow" errors. Also, it was casting the
3807 result into an int without checking.
3809 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3810 did not work. The code looks as if it has *never* worked, though it appears
3811 to have been documented since at least releast 1.62. I have made it work.
3813 35. The "timeout_DNS" error in retry rules, also documented since at least
3814 1.62, also never worked. As it isn't clear exactly what this means, and
3815 clearly it isn't a major issue, I have abolished the feature by treating it
3816 as "timeout", and writing a warning to the main and panic logs.
3818 36. The display of retry rules for -brt wasn't always showing the error code
3821 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3822 timeout_connect_A, timeout_connect_MX.
3824 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3825 to the empty sender.
3827 39. The daemon was not analysing the content of -oX till after it had closed
3828 stderr and disconnected from the controlling terminal. This meant that any
3829 syntax errors were only noted on the panic log, and the return code from
3830 the command was 0. By re-arranging the code a little, I've made the
3831 decoding happen first, so such errors now appear on stderr, and the return
3832 code is 1. However, the actual setting up of the sockets still happens in
3833 the disconnected process, so errors there are still only recorded on the
3836 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3837 connections (as happens on some IP stacks) was logged at start up time as
3838 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3839 it from "IPv6 and IPv4", which means that two separate sockets are being
3842 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3843 says whether AF_INET or AF_INET6 was passed as an argument.
3845 42. Exiwhat output was messed up when time zones were included in log
3848 43. Exiwhat now gives more information about the daemon's listening ports,
3849 and whether -tls-on-connect was used.
3851 44. The "port" option of the smtp transport is now expanded.
3853 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3854 silently ignored. Now an error message is written to the main and panic
3857 46. There's a new ACL modifier called "logwrite" which writes to a log file
3858 as soon as it is encountered.
3860 47. Added $local_user_uid and $local_user_gid at routing time.
3862 48. Exim crashed when trying to verify a sender address that was being
3865 49. Exim was recognizing only a space character after ".include". It now also
3866 recognizes a tab character.
3868 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3869 extracting the relevant information from the specification. The man page no
3870 longer contains scrambled data for the -d option, and I've added a section
3871 at the front about calling Exim under different names.
3873 51. Added "extra_headers" argument to the "mail" command in filter files.
3875 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3878 53. Installed eximstats 1.29.
3880 54. Added transport_filter_timeout as a generic transport option.
3882 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3883 Cc: header lines. This was required by RFC 822, but it not required by RFC
3886 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3887 incoming messages that did not have them. Now it does so only if the
3888 message originates locally, that is, if there is no associated remote host
3889 address. When Resent- header lines are present, this applies to the Resent-
3890 lines rather than the non-Resent- lines.
3892 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3893 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3895 58. Messages for configuration errors now include the name of the main
3896 configuration file - useful now that there may be more than one file in a
3897 list (.included file names were always shown).
3899 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3900 for those rare installations that do not start the daemon as root or run it
3901 setuid root. I've cut out the call to initgroups() if the daemon is not
3904 60. The Exim user and group can now be bound into the binary as text strings
3905 that are looked up at the start of Exim's processing.
3907 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3909 62. Added $mailstore_basename variable.
3911 63. Installed patch to sieve.c from Michael Haardt.
3913 64. When Exim failed to open the panic log after failing to open the main log,
3914 the original message it was trying to log was written to stderr and debug
3915 output, but if they were not available (the usual case in production), it
3916 was lost. Now it is written to syslog before the two lines that record the
3917 failures to open the logs.
3919 65. Users' Exim filters run in subprocesses under the user's uid. It is
3920 possible for a "deliver" command or an alias in a "personal" command to
3921 provoke an address rewrite. If logging of address rewriting is configured,
3922 this fails because the process is not running as root or exim. There may be
3923 a better way of dealing with this, but for the moment (because 4.30 needs
3924 to be released), I have disabled address rewrite logging when running a
3925 filter in a non-root, non-exim process.
3931 1. The buildconfig auxiliary program wasn't quoting the value set for
3932 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3933 not defined. This bug was masked in 4.22 by the effect that was fixed in
3936 2. Some messages that were rejected after a message id was allocated were
3937 shown as "incomplete" by exigrep. It no longer does this for messages that
3938 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3940 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3941 have allow_domain_literals set, the ID did not get logged in the <= line.
3942 Domain literals are now always recognized in Message-ID: header lines.
3944 4. The first argument for a ${extract expansion item is the key name or field
3945 number. Leading and trailing spaces in this item were not being ignored,
3946 causing some misleading effects.
3948 5. When deliver_drop_privilege was set, single queue runner processes started
3949 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3950 same command in the process it spins off) were not dropping privilege.
3952 6. When the daemon running as "exim" started a queue runner, it always
3953 re-executed Exim in the spun-off process. This is a waste of effort when
3954 deliver_drop_privilege is set. The new process now just calls the
3955 queue-runner function directly.
3961 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3964 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3965 ignored. Though the use of -oP was forcing the writing of a pid file, it
3966 was always written to the default place.
3968 3. If the message "no IP address found for host xxxx" is generated during
3969 incoming verification, it is now followed by identification of the incoming
3970 connection (so you can more easily find what provoked it).
3972 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3974 5. Added some features to "harden" Exim a bit more against certain attacks:
3976 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3977 be put in Local/Makefile. This is like the never_users runtime option,
3978 but it cannot be overridden. The default setting is "root".
3980 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3981 prefix string with which any file named in a -C command line option
3984 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3985 is retained for -C and -D only if the caller of Exim is root. Without
3986 it, the exim user may also use -C and -D and retain privilege.
3988 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3989 command line option is disabled.
3991 6. Macro names set by the -D option must start with an upper case letter, just
3992 like macro names defined in the configuration file.
3994 7. Added "dereference=" facility to LDAP.
3996 8. Two instances of the typo "uknown" in the source files are fixed.
3998 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3999 the Configure-Makefile script screwed up while processing it.
4001 10. Incorporated PCRE 4.4.
4003 11. The SMTP synchronization check was not operating right at the start of an
4004 SMTP session. For example, it could not catch a HELO sent before the client
4005 waited for the greeting. There is now a check for outstanding input at the
4006 point when the greeting is written. Because of the duplex, asynchronous
4007 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4008 way, but not yet received, when the check is performed.
4010 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4011 on TCP/IP sockets, because this apparently causes some broken clients to
4014 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4015 unchanged) from the Cygwin maintainer.
4017 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4018 was supported instead of testing for mbx. Effectively a typo.
4020 15. The spa authenticator server code was not checking that the input it
4021 received was valid base64.
4023 16. The debug output line for the "set" modifier in ACLs was not showing the
4024 name of the variable that was being set.
4026 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4028 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4029 lookup on its own. Something else was needed to trigger the lookup. For
4030 example, a match in host_lookup or the need for a host name in a host list.
4031 Now, if $sender_host_name is referenced and the host name has not yet been
4032 looked up, a lookup is performed. If the lookup fails, the variable remains
4033 empty, and $host_lookup_failed is set to "1".
4035 19. Added "eqi" as a case-independent comparison operator.
4037 20. The saslauthd authentication condition could segfault if neither service
4038 nor realm was specified.
4040 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4041 error message that was logged was misleading, and incoming SMTP
4042 connections were dropped. The message is now more accurate, and temporary
4043 errors are given to SMTP connections.
4045 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4046 (except RSET) with a 421 error, until QUIT is received. However, it was
4047 failing to send a response to QUIT.
4049 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4050 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4051 ACL is now run after the helo_try_verify_hosts code.
4053 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4054 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4055 case-independent, so other case variants are also recognized). Apparently
4056 some systems use these upper case variants.
4058 25. If more than two messages were waiting for the same host, and a transport
4059 filter was specified for the transport, Exim sent two messages over the
4060 same TCP/IP connection, and then failed with "socket operation on non-
4061 socket" when it tried to send the third.
4063 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4065 27. The extern definition of crypt16() in expand.c was not being excluded when
4066 the OS had its own crypt16() function.
4068 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4069 as a preferred synonym for return_size_limit, both as an option and as an
4072 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4074 30. Changed the default configuration ACL to relax the local part checking rule
4075 for addresses that are not in any local domains. For these addresses,
4076 slashes and pipe symbols are allowed within local parts, but the sequence
4077 /../ is explicitly forbidden.
4079 31. SPA server authentication was not clearing the challenge buffer before
4082 32. log_message in a "warn" ACL statement was writing to the reject log as
4083 well as to the main log, which contradicts the documentation and doesn't
4084 seem right (because no rejection is happening). So I have stopped it.
4086 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4087 However, I am unable to do any testing of this.
4089 34. Fixed an infelicity in the appendfile transport. When checking directories
4090 for a mailbox, to see if any needed to be created, it was accidentally
4091 using path names with one or more superfluous leading slashes; tracing
4092 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4094 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4095 contents are added to the log line that is written for every discarded
4096 recipient. (Previously a log_message setting was ignored.)
4098 36. The ${quote: operator now quotes the string if it is empty.
4100 37. The install script runs exim in order to find its version number. If for
4101 some reason other than non-existence or emptiness, which it checks, it
4102 could not run './exim', it was installing it with an empty version number,
4103 i.e. as "exim-". This error state is now caught, and the installation is
4106 38. An argument was missing from the function that creates an error message
4107 when Exim fails to connect to the socket for saslauthd authentication.
4108 This could cause Exim to crash, or give a corrupted message.
4110 39. Added isip, isip4, and isip6 to ${if conditions.
4112 40. The ACL variables $acl_xx are now saved with the message, and can be
4113 accessed later in routers, transports, and filters.
4115 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4116 strings in the file are not string-expanded.
4118 42. If a MAIL command specified a SIZE value that was too large to fit into an
4119 int variable, the check against message_size_limit failed. Such values are
4120 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4121 day this will have to be increased, but I don't think I want to be around
4122 when emails are that large.
4129 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4130 iconv() is not standard in FreeBSD.
4132 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4133 IPv6 enabled. The observed symptom was a segmentation fault on return from
4134 the function os_common_find_running_interfaces() in src/os.c.
4136 3. In the check_special_case() function in daemon.c I had used "errno" as an
4137 argument name, which causes warnings on some systems. This was basically a
4138 typo, since it was named "eno" in the comments!
4140 4. The code that waits for the clock to tick (at a resolution of some fraction
4141 of a second) so as to ensure message-id uniqueness was always waiting for
4142 at least one whole tick, when it could have waited for less. [This is
4143 almost certainly not relevant at current processor speeds, where it is
4144 unlikely to ever wait at all. But we try to future-proof.]
4146 5. The function that sleeps for a time interval that includes fractions of a
4147 second contained a race. It did not block SIGALRM between setting the
4148 timer, and suspending (a couple of lines later). If the interval was short
4149 and the sigsuspend() was delayed until after it had expired, the suspension
4150 never ended. On busy systems this could lead to processes getting stuck for
4153 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4154 process, before it forks any delivery processes. The open lookup caching
4155 mechanism meant that the open file or database connection was passed into
4156 the delivery process. The problem was that delivery processes always tidy
4157 up cached lookup data. This could cause a problem for the next delivery
4158 process started by the queue runner, because the external queue runner
4159 process does not know about the closure. So the next delivery process
4160 still has data in the lookup cache. In the case of a file lookup, there was
4161 no problem because closing a file descriptor in a subprocess doesn't affect
4162 the parent. However, if the lookup was caching a connection to a database,
4163 the connection was closed, and the second delivery process was likely to
4164 see errors such as "PGSQL: query failed: server closed the connection
4165 unexpectedly". The problem has been fixed by closing all cached lookups
4166 in a queue runner before running a delivery process.
4168 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4169 seem to have the "const" qualifier which it has on other OS. I've
4172 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4173 *of the same type* (client or server) with the same public name that an
4174 error should be diagnosed.
4176 9. When Exim looked up a host name for an IP address, but failed to find the
4177 original IP address when looking up the host name (a safety check), it
4178 output the message "<ip address> does not match any IP for NULL", which was
4179 confusing, to say the least. The bug was that the host name should have
4180 appeared instead of "NULL".
4182 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4183 user that is built into the binary, and the -C or -D options is used, root
4184 privilege is dropped before the configuration file is read. In addition,
4185 logging is switched to stderr instead of the normal log files. If the
4186 configuration then re-defines the Exim user, the unprivileged environment
4187 is probably not what is expected, so Exim logs a panic warning message (but
4190 However, if deliver_drop_privilege is set, the unprivileged state may well
4191 be exactly what is intended, so the warning has been cut out in that case,
4192 and Exim is allowed to try to write to its normal log files.
4198 1. smtp_return_error_details was not giving details for temporary sender
4199 or receiver verification errors.
4201 2. Diagnose a configuration error if two authenticators have the same public
4204 3. Exim used not to create the message log file for a message until the first
4205 delivery attempt. This could be confusing when incoming messages were held
4206 for policy or load reasons. The message log file is now created at the time
4207 the message is received, and an initial "Received" line is written to it.
4209 4. The automatically generated man page for command line options had a minor
4210 bug that caused no ill effects; however, a more serious problem was that
4211 the procedure for building the man page automatically didn't always
4212 operate. Consequently, release 4.20 contains an out-of-date version. This
4213 shouldn't happen again.
4215 5. When building Exim with embedded Perl support, the script that builds the
4216 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4217 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4219 6. The freeze_tell option was not being used for messages that were frozen on
4220 arrival, either by an ACL or by local_scan().
4222 7. Added the smtp_incomplete_transaction log selector.
4224 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4225 was accepting AUTH without a new EHLO.
4227 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4228 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4231 10. Logging of TCP/IP connections (when configured) now happens in the main
4232 daemon process instead of the child process, so that the TCP/IP connection
4233 count is more accurate (but it can never be perfect).
4235 11. The use of "drop" in a nested ACL was not being handled correctly in the
4236 outer ACL. Now, if condition failure induced by the nested "drop" causes
4237 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4238 or "require"), the connection is dropped.
4240 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4241 that yield "discard" can now be used with an "accept" or a "discard" verb,
4242 but an error is generated for any others (because I can't see a useful way
4243 to define what should happen).
4245 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4246 are now processed in the same way as lines in the Exim configuration file.
4247 In particular, continuation lines are supported.
4249 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4251 15. Added -ti meaning -t -i.
4253 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4254 domains, and warn by logging if others are found.
4256 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4257 returned from GIFCONF when Exim is trying to find the list of interfaces on
4258 a host. The code in os.c has been modified to copy each ifreq to an aligned
4259 structure in all cases.
4261 Also, in some cases, the returned ifreq's were being copied to a 'struct
4262 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4263 means the last couple of bytes of an IPv6 address could be chopped if the
4264 ifreq contained only a normal sockaddr (14 bytes storage).
4266 18. Named domain lists were not supported in the hosts_treat_as_local option.
4267 An entry such as +xxxx was not recognized, and was treated as a literal
4270 19. Ensure that header lines added by a DATA ACL are included in the reject log
4271 if the ACL subsequently rejects the message.
4273 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4274 MD5 (which is deprecated).
4276 21. When testing a filter file using -bf, Exim was writing a message when it
4277 took the sender from a "From " line in the message, but it was not doing so
4278 when it took $return_path from a Return-Path: header line. It now does.
4280 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4281 with a valid header line field name (a series of printing characters
4282 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4284 23. Changed "disc" in the source to "disk" to conform to the documentation and
4285 the book and for uniformity.
4287 24. Ignore Sendmail's -Ooption=value command line item.
4289 25. When execve() failed while trying to run a command in a pipe transport,
4290 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4291 could be confused with a return value of 69 from the command itself. This
4292 has been changed to 127, the value the shell returns if it is asked to run
4293 a non-existent command. The wording for the related log line suggests a
4294 non-existent command as the problem.
4296 26. If received_header_text expands to an empty string, do not add a Received:
4297 header line to the message. (Well, it adds a token one on the spool, but
4298 marks it "old" so that it doesn't get used or transmitted.)
4300 27. Installed eximstats 1.28 (addition of -nt option).
4302 28. There was no check for failure on the call to getsockname() in the daemon
4303 code. This can fail if there is a shortage of resources on the system, with
4304 ENOMEM, for example. A temporary error is now given on failure.
4306 29. Contrary to the C standard, it seems that in some environments, the
4307 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4308 program. Exim now does this explicitly; it affects the formatting of
4309 timestamps using strftime().
4311 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4312 complaints. I've now initialized all the variables, to avoid this.
4314 32. Header lines added by a system filter were not being "seen" during
4315 transport-time rewrites.
4317 33. The info_callback() function passed to OpenSSL is set up with type void
4318 (*)(SSL *, int, int), as described somewhere. However, when calling the
4319 function (actually a macro) that sets it up, the type void(*)() is
4320 expected. I've put in a cast to prevent warnings from picky compilers.
4322 34. If a DNS black list lookup found a CNAME record, but there were no A
4323 records associated with the domain it pointed at, Exim crashed.
4325 35. If a DNS black list lookup returned more than one A record, Exim ignored
4326 all but the first. It now scans all returned addresses if a particular IP
4327 value is being sought. In this situation, the contents of the
4328 $dnslist_value variable are a list of all the addresses, separated by a
4331 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4332 to accept a host name and all its aliases if the forward lookup for any of
4333 them yielded the IP address of the incoming connection. Now it accepts only
4334 those names whose forward lookup yields the correct IP address. Any other
4335 names are discarded. This closes a loophole whereby a rogue DNS
4336 administrator could create reverse DNS records to break through a
4337 wildcarded host restriction in an ACL.
4339 37. If a user filter or a system filter that ran in a subprocess used any of
4340 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4341 the wrong values were passed to the pipe command ($thisaddress had the
4342 value of $0, $0 had the value of $1, etc). This bug was introduced by
4343 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4345 38. Improved the line breaking for long SMTP error messages from ACLs.
4346 Previously, if there was no break point between 40 and 75 characters, Exim
4347 left the rest of the message alone. Two changes have been made: (a) I've
4348 reduced the minimum length to 35 characters; (b) if it can't find a break
4349 point between 35 and 75 characters, it looks ahead and uses the first one
4350 that it finds. This may give the occasional overlong line, but at least the
4351 remaining text gets split now.
4353 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4354 file descriptors might be low, and that setting 1000 would always raise it.
4355 It turns out that in some environments, the limit is already over 1000 and
4356 that lowering it causes trouble. So now Exim takes care not to decrease it.
4358 40. When delivering a message, the value of $return_path is set to $sender_
4359 address at the start of routing (routers may change the value). By an
4360 oversight, this default was not being set up when an address was tested by
4361 -bt or -bv, which affected the outcome if any router or filter referred to
4364 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4365 log only when "message" or "log_message" are set. However, if one of the
4366 conditions was an address verification, or a call to a nested ACL, the
4367 messages generated by the underlying test were being passed through. This
4368 no longer happens. The underlying message is available in $acl_verify_
4369 message for both "message" and "log_message" expansions, so it can be
4370 passed through if needed.
4372 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4373 new expansion $bh_ to give the encoded byte string without charset
4374 translation. Translation happens only if iconv() is available; HAVE_ICONV
4375 indicates this at build time. HEADERS_CHARSET gives the charset to
4376 translate to; headers_charset can change it in the configuration, and
4377 "headers charset" can change it in an individual filter file.
4379 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4380 that creates RFC 2047 encoded "words" labels them as that charset instead
4381 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4382 expansion operator; (ii) when Exim creates a From: line for a local
4383 message; (iii) when a header line is rewritten to include a "phrase" part.
4385 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4386 buggy, causing it to skip the first lines of messages whose message ID
4387 ended in 'D'. This would not have bitten before Exim release 4.14, because
4388 message IDs were unlikely to end in 'D' before then. The effect was to have
4389 incorrect size information for certain domains.
4391 45. #include "config.h" was missing at the start of the crypt16.c module. This
4392 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4395 46. If there was a timeout during a "random" callout check, Exim treated it as
4396 a failure of the random address, and carried on sending RSET and the real
4397 address. If the delay was just some slowness somewhere, the response to the
4398 original RCPT would be taken as a response to RSET and so on, causing
4399 mayhem of various kinds.
4401 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4402 when I implemented it. It didn't allow for the fact that some option values
4403 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4404 the right test for positive values.
4406 48. Domain names in DNS records are case-independent. Exim always looks them up
4407 in lower case. Some resolvers return domain names in exactly the case they
4408 appear in the zone file, that is, they may contain uppercase letters. Not
4409 all resolvers do this - some return always lower case. Exim was treating a
4410 change of case by a resolver as a change of domain, similar to a widening
4411 of a domain abbreviation. This triggered its re-routing code and so it was
4412 trying to route what was effectively the same domain again. This normally
4413 caused routing to fail (because the router wouldn't handle the domain
4414 twice). Now Exim checks for this case specially, and just changes the
4415 casing of the domain that it ultimately uses when it transmits the message
4418 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4421 50. If a filter generated a file delivery with a non-absolute name (possible if
4422 no home directory exists for the router), the forbid_file option was not
4425 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4426 the existing equality matching.
4428 52. Exim was using ints instead of ino_t variables in some places where it was
4429 dealing with inode numbers.
4431 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4432 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4433 called TMPDIR, and if it finds it is different, it changes its value.
4435 54. The smtp_printf() function is now made available to local_scan() so
4436 additional output lines can be written before returning. There is also an
4437 smtp_fflush() function to enable the detection of a dropped connection.
4438 The variables smtp_input and smtp_batched_input are exported to
4441 55. Changed the default runtime configuration: the message "Unknown user"
4442 has been removed from the ACL, and instead placed on the localuser router,
4443 using the cannot_route_message feature. This means that any verification
4444 failures that generate their own messages won't get overridden. Similarly,
4445 the "Unrouteable address" message that was in the ACL for unverifiable
4446 relay addresses has also been removed.
4448 56. Added hosts_avoid_esmtp to the smtp transport.
4450 57. The exicyclog script was not checking for the esoteric option
4451 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4452 will work only if exicyclog is run under the appropriate euid.
4454 58. Following a discussion on the list, the rules by which Exim recognises line
4455 endings on incoming messages have been changed. The -dropcr and drop_cr
4456 options are now no-ops, retained only for backwards compatibility. The
4457 following line terminators are recognized: LF CRLF CR. However, special
4458 processing applies to CR:
4460 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4461 nor a local message in the state where . is a terminator.
4463 (ii) If a bare CR is encountered in a header line, an extra space is added
4464 after the line terminator so as not to end the header. The reasoning
4465 behind this is that bare CRs in header lines are most likely either
4466 to be mistakes, or people trying to play silly games.
4468 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4469 was being incorrectly given as 18 bytes larger than it should have been.
4470 This is a VOB (very old bug).
4472 60. This may never have affected anything current, but just in case it has:
4473 When the local host is found other than at the start of a list of hosts,
4474 the local host, those with the same MX, and any that follow, are discarded.
4475 When the list in question was part of a longer list of hosts, the following
4476 hosts (not currently being processed) were also being discarded. This no
4477 longer happens. I'm not sure if this situation could ever has previously
4480 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4483 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4484 header. This is something that is recommended in a new Internet Draft, and
4485 is something that is documented as being done by Sendmail. There are two
4486 possible values. For messages generated by the autoreply transport, Exim
4489 Auto-Submitted: auto-replied
4491 whereas for all other generated messages (e.g. bounces) it adds
4493 Auto-Submitted: auto-generated
4495 63. The "personal" condition in filters now includes a test for the
4496 Auto-Submitted: header. If it contains the string "auto-" the message it
4497 not considered personal.
4499 64. Added rcpt_include_affixes as a generic transport option.
4501 65. Added queue_only_override (default true).
4503 66. Added the syslog_duplication option.
4505 67. If what should have been the first header line of a message consisted of
4506 a space followed by a colon, Exim was mis-interpreting it as a header line.
4507 It isn't of course - it is syntactically invalid and should therefore be
4508 treated as the start of the message body. The misbehaviour could have
4509 caused a number of strange effects, including loss of data in subsequent
4510 header lines, and spool format errors.
4512 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4513 client host had authenticated. This control can now be exercised by an ACL
4514 for more flexibility.
4516 69. By default, callouts do not happen when testing with -bh. There is now a
4517 variant, -bhc, which does actually run the callout code, including
4518 consulting and updating the callout cache.
4520 70. Added support for saslauthd authentication, courtesy of Alexander
4523 71. If statvfs() failed on the spool or log directories while checking their
4524 size for availability, Exim confusingly gave the error "space shortage".
4525 Furthermore, in debugging mode it crashed with a floating point exception.
4526 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4527 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4528 serious problem, Exim now writes to the main and panic logs when this
4529 happens, with details of the failure. It then refuses to accept the
4530 incoming message, giving the message "spool directory problem" or "log
4531 directory problem" with a 421 code for SMTP messages.
4533 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4534 0, 1, and 2 exist, because some OS complain for execs without them (see
4535 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4536 descriptors. However, the code omitted to check that the open succeeded,
4537 causing mysterious errors if for some reason the permissions on /dev/null
4538 got screwed. Now Exim writes a message to the main and panic logs, and
4539 bombs out if it can't open /dev/null.
4541 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4542 interact so that it is all more flexible. It is supposed to remain
4543 backwards compatible. Also added extra_local_interfaces.
4545 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4546 to bomb out with an assertion failure - to the client this appears as a
4547 connection drop. This problem occurs in the part of the code that was taken
4548 from the Samba project. Fortunately, the assertion is in a very simple
4549 function, so I have fixed this by reproducing the function inline in the
4550 one place where it is called, and arranging for authentication to fail
4551 instead of killing the process with assert().
4553 75. The SPA client code was not working when the server requested OEM rather
4554 than Unicode encoding.
4556 76. Added code to make require_files with a specific uid setting more usable in
4557 the case where statting the file as root fails - usually a non-root-mounted
4558 NFS file system. When this happens and the failure is EACCES, Exim now
4559 forks a subprocess and does the per-uid checking as the relevant uid.
4561 77. Added process_log_path.
4563 78. If log_file_path was not explicitly set, a setting of check_log_space or
4564 check_log_inodes was ignored.
4566 79. If a space check for the spool or log partitions fails, the incident is now
4567 logged. Of course, in the latter case the data may get lost...
4569 80. Added the %p formatting code to string_format() so that it can be used to
4570 print addresses in debug_print(). Adjusted all the address printing in the
4571 debugging in store.c to use %p rather than %d.
4573 81. There was a concern that a line of code in smtp_in.c could overflow a
4574 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4575 initially expressed, the concern was not well-founded, because trailing
4576 spaces are removed early. However, if the trailing spaces were followed by
4577 a NULL, they did not get removed, so the overflow was possible. Two fixes
4580 (a) I re-wrote the offending code in a cleaner fashion.
4581 (b) If an incoming SMTP command contains a NULL character, it is rejected
4584 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4585 runs initgroups(), so that if the Exim user is in any additional groups,
4586 they will be used during message reception.
4592 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob