1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.465 2007/01/30 15:10:58 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
74 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
75 fixed by 4.65/MH/01 (is this a record?) are fixed:
77 (i) An empty string was always treated as zero by the numeric comparison
78 operators. This behaviour has been restored.
80 (ii) It is documented that the numeric comparison operators always treat
81 their arguments as decimal numbers. This was broken in that numbers
82 starting with 0 were being interpreted as octal.
84 While fixing these problems I realized that there was another issue that
85 hadn't been noticed. Values of message_size_limit (both the global option
86 and the transport option) were treated as octal if they started with 0.
87 The documentation was vague. These values are now always treated as
88 decimal, and I will make that clear in the documentation.
94 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
95 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
98 MH/01 Don't check that the operands of numeric comparison operators are
99 integers when their expansion is in "skipping" mode (fixes bug
100 introduced by 4.64-PH/07).
102 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
103 child addresses, Exim now panics and dies. Previously, because the count
104 is held in a short int, deliveries were likely to be lost. As such a
105 large number of recipients for a single message is ridiculous
106 (performance will be very, very poor), I have chosen to impose a limit
107 rather than extend the field.
113 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
114 leftover -K file (the existence of which was triggered by #402).
115 While we were at it, introduced process PID as part of the -K
116 filename. This should rule out race conditions when creating
119 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
120 processing considerably. Previous code took too long for large mails,
121 triggering a timeout which in turn triggers #401.
123 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
124 in the DK code in transports.c. sendfile() is not really portable,
125 hence the _LINUX specificness.
127 TF/01 In the add_headers option to the mail command in an Exim filter,
128 there was a bug that Exim would claim a syntax error in any
129 header after the first one which had an odd number of characters
132 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
133 callout verification, Exim cached a "reject" for the entire domain. This
134 is correct for most verifications, but it is not correct for a recipient
135 verification with use_sender or use_postmaster set, because in that case
136 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
137 case of MAIL FROM:<> rejection from other early rejections (e.g.
138 rejection of HELO). When verifying a recipient using a non-null MAIL
139 address, the cache is ignored if it shows MAIL FROM:<> rejection.
140 Whatever the result of the callout, the value of the domain cache is
141 left unchanged (for any other kind of callout, getting as far as trying
142 RCPT means that the domain itself is ok).
144 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
147 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
148 manifest itself as EPIPE rather than ECONNECT. When tidying away a
149 session, the daemon ignores ECONNECT errors and logs others; it now
150 ignores EPIPE as well.
152 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
153 (quoted-printable decoding).
155 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
156 later the small subsequent patch to fix an introduced bug.
158 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
160 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
162 PH/08 An error is now given if message_size_limit is specified negative.
164 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
165 to be given (somewhat) arbitrary names.
167 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
170 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
171 miscellaneous code fixes
173 PH/10 Added the log_reject_target ACL modifier to specify where to log
176 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
177 hostname. This is wrong, because it relates to the incoming message (and
178 probably the interface on which it is arriving) and not to the outgoing
179 callout (which could be using a different interface). This has been
180 changed to use the value of the helo_data option from the smtp transport
181 instead - this is what is used when a message is actually being sent. If
182 there is no remote transport (possible with a router that sets up host
183 addresses), $smtp_active_hostname is used.
185 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
186 tweaks were necessary in order to get it to work (see also 21 below):
187 (a) The code assumed that strncpy() returns a negative number on buffer
188 overflow, which isn't the case. Replaced with Exim's string_format()
190 (b) There were several signed/unsigned issues. I just did the minimum
191 hacking in of casts. There is scope for a larger refactoring.
192 (c) The code used strcasecmp() which is not a standard C function.
193 Replaced with Exim's strcmpic() function.
194 (d) The code set only $1; it now sets $auth1 as well.
195 (e) A simple test gave the error "authentication client didn't specify
196 service in request". It would seem that Dovecot has changed its
197 interface. Fortunately there's a specification; I followed it and
198 changed what the client sends and it appears to be working now.
200 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
203 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
204 address A is aliased to B and C, where B exists and C does not. Without
205 -v the output is "A verified" because verification stops after a
206 successful redirection if more than one address is generated. However,
207 with -v the child addresses are also verified. Exim was outputting "A
208 failed to verify" and then showing the successful verification for C,
209 with its parentage. It now outputs "B failed to verify", showing B's
210 parentage before showing the successful verification of C.
212 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
213 look up a TXT record in a specific list after matching in a combined
216 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
217 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
218 they consult the DNS. I had assumed they would set it the way they
219 wanted; and indeed my experiments on Linux seem to show that in some
220 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
221 To be on the safe side, however, I have now made the interface to
222 host_find_byname() similar to host_find_bydns(), with an argument
223 containing the DNS resolver options. The host_find_byname() function now
224 sets these options at its start, just as host_find_bydns() does. The smtp
225 transport options dns_qualify_single and dns_search_parents are passed to
226 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
227 of host_find_byname() use the default settings of RES_DEFNAMES
228 (qualify_single) but not RES_DNSRCH (search_parents).
230 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
231 spool_read_header() do less string testing, by means of a preliminary
232 switch on the second character of optional "-foo" lines. (This is
233 overdue, caused by the large number of possibilities that now exist.
234 Originally there were few.) While I was there, I also converted the
235 str(n)cmp tests so they don't re-test the leading "-" and the first
236 character, in the hope this might squeeze out yet more improvement.
238 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
239 flag allowing group syntax was set by the header_syntax check but not
240 turned off, possible causing trouble later; (2) The flag was not being
241 set at all for the header_verify test, causing "group"-style headers to
242 be rejected. I have now set it in this case, and also caused header_
243 verify to ignore an empty address taken from a group. While doing this, I
244 came across some other cases where the code for allowing group syntax
245 while scanning a header line wasn't quite right (mostly, not resetting
246 the flag correctly in the right place). These bugs could have caused
247 trouble for malformed header lines. I hope it is now all correct.
249 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
250 with the "reply" argument non-NULL. The code, however (which originally
251 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
252 but it didn't always do it. This confused somebody who was copying the
253 code for some other use. I have removed all the tests.
255 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
256 feature that was used to support insecure browsers during the U.S. crypto
257 embargo. It requires special client support, and Exim is probably the
258 only MTA that supported it -- and would never use it because real RSA is
259 always available. This code has been removed, because it had the bad
260 effect of slowing Exim down by computing (never used) parameters for the
261 RSA_EXPORT functionality.
263 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
264 authenticator to fail if there's a tab character in the incoming data
265 (there should never be unless someone is messing about, as it's supposed
266 to be base64-encoded). Also added, on Timo's advice, the "secured" option
267 if the connection is using TLS or if the remote IP is the same as the
268 local IP, and the "valid-client-cert option" if a client certificate has
271 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
272 authenticators. This can be used for authorization after authentication
273 succeeds. (In the case of plaintext, it servers for both authentication
276 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
277 if any retry times were supplied.
279 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
280 connection, where there is no client IP address to check. In this
281 situation, the verify now always succeeds.
283 PH/25 Applied John Jetmore's -Mset patch.
285 PH/26 Added -bem to be like -Mset, but loading a message from a file.
287 PH/27 In a string expansion for a processed (not raw) header when multiple
288 headers of the same name were present, leading whitespace was being
289 removed from all of them, but trailing whitespace was being removed only
290 from the last one. Now trailing whitespace is removed from each header
291 before concatenation. Completely empty headers in a concatenation (as
294 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
295 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
297 PH/29 [Removed. This was a change that I later backed out, and forgot to
298 correct the ChangeLog entry (that I had efficiently created) before
299 committing the later change.]
301 PH/30 Exim was sometimes attempting to deliver messages that had suffered
302 address errors (4xx response to RCPT) over the same connection as other
303 messages routed to the same hosts. Such deliveries are always "forced",
304 so retry times are not inspected. This resulted in far too many retries
305 for the affected addresses. The effect occurred only when there were more
306 hosts than the hosts_max_try setting in the smtp transport when it had
307 the 4xx errors. Those hosts that it had tried were not added to the list
308 of hosts for which the message was waiting, so if all were tried, there
309 was no problem. Two fixes have been applied:
311 (i) If there are any address or message errors in an SMTP delivery, none
312 of the hosts (tried or untried) are now added to the list of hosts
313 for which the message is waiting, so the message should not be a
314 candidate for sending over the same connection that was used for a
315 successful delivery of some other message. This seems entirely
316 reasonable: after all the message is NOT "waiting for some host".
317 This is so "obvious" that I'm not sure why it wasn't done
318 previously. Hope I haven't missed anything, but it can't do any
319 harm, as the worst effect is to miss an optimization.
321 (ii) If, despite (i), such a delivery is accidentally attempted, the
322 routing retry time is respected, so at least it doesn't keep
323 hammering the server.
325 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
326 in ${readsocket because some servers need this prod.
328 PH/32 Added some extra debug output when updating a wait-xxx database.
330 PH/33 The hint "could be header name not terminated by colon", which has been
331 given for certain expansion errors for a long time, was not being given
332 for the ${if def:h_colon_omitted{... case.
334 PH/34 The spec says: "With one important exception, whenever a domain list is
335 being scanned, $domain contains the subject domain." There was at least
336 one case where this was not true.
338 PH/35 The error "getsockname() failed: connection reset by peer" was being
339 written to the panic log as well as the main log, but it isn't really
340 panic-worthy as it just means the connection died rather early on. I have
341 removed the panic log writing for the ECONNRESET error when getsockname()
344 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
345 runs only) independently of the message's sender address. This meant
346 that, if the 4xx error was in fact related to the sender, a different
347 message to the same recipient with a different sender could confuse
348 things. In particualar, this can happen when sending to a greylisting
349 server, but other circumstances could also provoke similar problems.
350 I have changed the default so that the retry time for these errors is now
351 based a combination of the sender and recipient addresses. This change
352 can be overridden by setting address_retry_include_sender=false in the
355 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
356 remote server are returned as part of bounce messages. This was not
357 happening for LMTP over a pipe (the lmtp transport), but now it is the
358 same for both kinds of LMTP.
360 PH/38 Despite being documented as not happening, Exim was rewriting addresses
361 in header lines that were in fact CNAMEs. This is no longer the case.
363 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
364 and queue runs started by the daemon processed all messages. This has
365 been fixed so that -R and -S can now usefully be given with -q<time>.
367 PH/40 Import PCRE release 6.7 (fixes some bugs).
369 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
371 PH/42 Give an error if -q is specified more than once.
373 PH/43 Renamed the variables $interface_address and $interface_port as
374 $received_ip_address and $received_port, to make it clear that these
375 values apply to message reception, and not to the outgoing interface when
376 a message is delivered. (The old names remain recognized, of course.)
378 PH/44 There was no timeout on the connect() call when using a Unix domain
379 socket in the ${readsocket expansion. There now is.
381 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
382 be meaningful with "accept".
384 SC/01 Eximstats V1.43
385 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
387 SC/02 Eximstats V1.44
388 Use a glob alias rather than an array ref in the generated
389 parser. This improves both readability and performance.
391 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
392 Collect SpamAssassin and rejection statistics.
393 Don't display local sender or destination tables unless
394 there is data to show.
395 Added average volumes into the top table text output.
397 SC/04 Eximstats V1.46
398 Collect data on the number of addresses (recipients)
399 as well as the number of messages.
401 SC/05 Eximstats V1.47
402 Added 'Message too big' to the list of mail rejection
403 reasons (thanks to Marco Gaiarin).
405 SC/06 Eximstats V1.48
406 Mainlog lines which have GMT offsets and are too short to
407 have a flag are now skipped.
409 SC/07 Eximstats V1.49 (Alain Williams)
410 Added the -emptyok flag.
412 SC/08 Eximstats V1.50
413 Fixes for obtaining the IP address from reject messages.
415 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
416 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
417 whitesspace changes from 4.64-PH/27
419 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
422 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
423 are found, allow negative numbers in numeric criteria)
425 JJ/06 exipick.20061117.2, added new $message_body_missing variable
427 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
428 to match changes made in 4.64-PH/43
430 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
432 PH/47 Put in an explicit test for a DNS lookup of an address record where the
433 "domain" is actually an IP address, and force a failure. This locks out
434 those revolvers/nameservers that support "A-for-A" lookups, in
435 contravention of the specifications.
437 PH/48 When a host name was looked up from an IP address, and the subsequent
438 forward lookup of the name timed out, the host name was left in
439 $sender_host_name, contrary to the specification.
441 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
442 restricted to single-key lookups, Exim was not diagnosing an error if
443 * or *@ was used with a query-style lookup.
445 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
447 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
448 long ago, but noone interested enough thought of it. Let's just say that
449 the "1.1" means that there are some new functions that weren't there at
450 some point in the past.
452 PH/51 Error processing for expansion failure of helo_data from an smtp
453 transport during callout processing was broken.
455 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
456 tested/used via the -bh/-bhc/-bs options.
458 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
459 bug, fixed in subsequent PCRE releases).
461 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
462 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
464 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
470 SC/01 Use a glob alias rather than an array ref in eximstats generated
471 parser. This improves both readability and performance.
473 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
474 Don't display local sender or destination tables in eximstats unless
475 there is data to show.
476 Added average volumes into the eximstats top table text output.
478 SC/03 Collect data on the number of addresses (recipients) as well
479 as the number of messages in eximstats.
481 TF/01 Correct an error in the documentation for the redirect router. Exim
482 does (usually) call initgroups() when daemonizing.
484 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
485 with consistent privilege compared to when running as a daemon.
487 TF/03 Note in the spec that $authenticated_id is not set for local
488 submissions from trusted users.
490 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
491 Thanks to Dean Brooks <dean@iglou.com> for the patch.
493 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
494 by adding some example configuration directives to the default
495 configuration file. A little bit of work is required to uncomment the
496 directives and define how usernames and passwords are checked, but
497 there is now a framework to start from.
499 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
500 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
501 without this. I don't know how relevant this is to other LDAP libraries.
503 PH/02 Add the verb name to the "unknown ACL verb" error.
505 PH/03 Magnus Holmgren's patch for filter_prepend_home.
507 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
509 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
510 directory not expanded when it should be if an expanded home directory
511 was set for the address (which is overridden by the transport).
513 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
516 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
517 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
518 because it is too late at that time, and has no effect.
520 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
521 security issue with \' (bugzilla #107). I could not use the
522 PQescapeStringConn() function, because it needs a PGconn value as one of
525 PH/08 When testing addresses using -bt, indicate those final addresses that
526 are duplicates that would not cause an additional delivery. At least one
527 person was confused, thinking that -bt output corresponded to deliveries.
528 (Suppressing duplicates isn't a good idea as you lose the information
529 about possibly different redirections that led to the duplicates.)
531 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
532 systems where poll() doesn't work, in particular OS X.
534 PH/10 Added more information to debugging output for retry time not reached.
536 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
537 operations in malware.c.
539 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
542 PH/13 If write_rejectlog was set false when logging was sent to syslog with
543 syslog_duplication set false, log lines that would normally be written
544 both the the main log and to the reject log were not written to syslog at
547 PH/14 In the default configuration, change the use of "message" in ACL warn
548 statements to "add_header".
550 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
551 not followed by a command (e.g. "seen endif").
553 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
554 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
557 PH/17 Added extra conditions to the default value of delay_warning_condition
561 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
562 { match{$h_precedence:}{(?i)bulk|list|junk} } \
563 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
566 The Auto-Submitted: and various List- headers are standardised, whereas I
567 don't think Precedence: ever was.
569 PH/18 Refactored debugging code in route_finduser() to show more information,
570 in particular, the error code if getpwnam() issues one.
572 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
573 This is apparently needed in addition to the PH/07 change above to avoid
574 any possible encoding problems.
576 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
577 but not after initializing Perl.
579 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
580 output them only if debugging. By default they are written stderr,
581 apparently, which is not desirable.
583 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
586 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
589 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
591 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
592 authenticated or an ident call has been made. Suppress the default
593 values for $authenticated_id and $authenticated_sender (but permit -oMai
594 and -oMas) when testing with -bh.
596 PH/24 Re-jigged the order of the tests in the default configuration so that the
597 tests for valid domains and recipients precede the DNS black list and CSA
598 tests, on the grounds that those ones are more expensive.
600 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
601 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
602 This bug exists in every version of Exim that I still have, right back to
605 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
606 However, an attempt to turn on case-sensitivity in a regex key by
607 including (?-i) didn't work because the subject string was already
608 lowercased, and the effects were non-intuitive. It turns out that a
609 one-line patch can be used to allow (?-i) to work as expected.
615 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
616 other effects) broke the use of negated acl sub-conditions.
618 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
621 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
622 "Deny" causes Exim to reject the incoming connection with a 554 error.
623 Unfortunately, if there is a major crisis, such as a disk failure,
624 tcp-wrappers gives "deny", whereas what one would like would be some
625 kind of temporary error. A kludge has been added to help with this.
626 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
627 554 error is used if errno is still zero or contains ENOENT (which occurs
628 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
631 PH/03 Add -lutil to the default FreeBSD LIBS setting.
633 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
634 errors. Otherwise a message that provokes a temporary error (when other
635 messages do not) can cause a whole host to time out.
637 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
638 addresses were routed directly to files or pipes from a redirect router.
639 File deliveries just didn't batch; pipe deliveries might have suffered
642 PH/06 A failure to get a lock for a hints database would erroneously always say
643 "Failed to get write lock", even when it was really a read lock.
645 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
646 of 0600. This has been changed to use the value of the lockfile_mode
647 option (which defaults to 0600).
649 PH/08 Applied small patch from the Sieve maintainer.
651 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
652 folder from quota calculations, a direct delivery into this folder messed
653 up the contents of the maildirsize file. This was because the regex was
654 used only to exclude .Trash (or whatever) when the size of the mailbox
655 was calculated. There was no check that a delivery was happening into an
656 excluded directory. This bug has been fixed by ignoring all quota
657 processing for deliveries into excluded directories.
659 PH/10 Added the maildirfolder_create_regex option to appendfile.
665 PH/01 The code for finding all the local interface addresses on a FreeBSD
666 system running IPv6 was broken. This may well have applied to all BSD
667 systems, as well as to others that have similar system calls. The broken
668 code found IPv4 interfaces correctly, but gave incorrect values for the
669 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
670 that it would not match correctly against @[] and not recognize the IPv6
673 PH/02 The ipliteral router was not recognizing addresses of the form user@
674 [ipv6:....] because it didn't know about the "ipv6:" prefix.
676 PH/03 Added disable_ipv6.
678 PH/04 Changed $reply_address to use the raw form of the headers instead of the
679 decoded form, because it is most often used to construct To: headers
680 lines in autoreplies, and the decoded form may well be syntactically
681 invalid. However, $reply_address has leading white space removed, and all
682 newlines turned into spaces so that the autoreply transport does not
685 PH/05 If group was specified without a user on a router, and no group or user
686 was specified on a transport, the group from the router was ignored.
688 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
689 for visible compile-time settings that can be used to change these
690 numbers, for those that want even more. Backwards compatibility with old
691 spool files has been maintained. However, going back to a previous Exim
692 release will lost any variables that are in spool files.
694 PH/07 Two small changes when running in the test harness: increase delay when
695 passing a TCP/IP connection to a new process, in case the original
696 process has to generate a bounce, and remove special handling of
697 127.0.0.2 (sic), which is no longer necessary.
699 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
700 be the same on different OS.
702 PH/09 Moved a debug statement in filter processing to avoid a race problem when
705 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
706 whether --show-vars was specified or not
708 JJ/02 exipick: Added support for new ACL variable spool format introduced
711 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
712 syntactically invalid From: or Reply-to: line, and a filter used this to
713 generate an autoreply, and therefore failed to obtain an address for the
714 autoreply, Exim could try to deliver to a non-existent relative file
715 name, causing unrelated and misleading errors. What now happens is that
716 it logs this as a hard delivery error, but does not attempt to create a
719 PH/11 The exinext utility has a -C option for testing purposes, but although
720 the given file was scanned by exinext itself; it wasn't being passed on
721 when Exim was called.
723 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
724 an end-of-file indication when reading a command response.
726 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
727 compiled. In many other places in Exim, IPv6 addresses are always
728 recognized, so I have changed this. It also means that IPv4 domain
729 literals of the form [IPV4:n.n.n.n] are now always recognized.
731 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
732 used if the router is not running as root, for example, when verifying at
733 ACL time, or when using -bh. The debugging output from this situation was
734 non-existent - all you got was a failure to exec. I have made two
737 (a) Failures to set uid/gid, the current directory, or a process leader
738 in a subprocess such as that created by queryprogram now generate
739 suitable debugging ouput when -d is set.
741 (b) The queryprogram router detects when it is not running as root,
742 outputs suitable debugging information if -d is set, and then runs
743 the subprocess without attempting to change uid/gid.
745 PH/15 Minor change to Makefile for building test_host (undocumented testing
748 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
749 additional section of a DNS packet that returns MX or SRV records.
750 Instead, it always explicitly searches for A/AAAA records. This avoids
751 major problems that occur when a DNS server includes only records of one
752 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
753 fixed another bug: if SRV records were looked up and the corresponding
754 address records were *not* found in the additional section, the port
755 values from the SRV records were lost.
757 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
758 using the correct key (the original address) when searching the retry
759 rules in order to find which one to use for generating the retry hint.
761 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
762 adding the default one. Similarly, if it contains a Reply-To: header, the
763 errors_reply_to option, if set, is not used.
765 PH/19 When calculating a retry time, Exim used to measure the "time since
766 failure" by looking at the "first failed" field in the retry record. Now
767 it does not use this if it is later than than the arrival time of the
768 message. Instead it uses the arrival time. This makes for better
769 behaviour in cases where some deliveries succeed, thus re-setting the
770 "first failed" field. An example is a quota failure for a huge message
771 when small messages continue to be delivered. Without this change, the
772 "time since failure" will always be short, possible causing more frequent
773 delivery attempts for the huge message than are intended.
774 [Note: This change was subsequently modified - see PH/04 for 4.62.]
776 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
777 $1, $2, $3) because the numerical variables can be reset during some
778 expansion items (e.g. "match"), thereby losing the authentication data.
780 PH/21 Make -bV show the size of off_t variables so that the test suite can
781 decide whether to run tests for quotas > 2G.
783 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
784 mailbox_size, and mailbox_filecount in the appendfile transport. If a
785 filecount value is greater than 2G or if a quota value is greater than 2G
786 on a system where the size of off_t is not greater than 4, a panic error
789 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
790 never match. The debug and -bh output now contains an explicit error
791 message indicating a malformed IPv4 address or mask.
793 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
794 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
797 PH/25 Do not write to syslog when running in the test harness. The only
798 occasion when this arises is a failure to open the main or panic logs
799 (for which there is an explicit test).
801 PH/26 Added the /no_tell option to "control=freeze".
803 PH/27 If a host name lookup failed very early in a connection, for example, if
804 the IP address matched host_lookup and the reverse lookup yielded a name
805 that did not have a forward lookup, an error message of the form "no IP
806 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
807 could be logged. Now it outputs the IP address instead of "NULL".
809 PH/28 An enabling patch from MH: add new function child_open_exim2() which
810 allows the sender and the authenticated sender to be set when
811 submitting a message from within Exim. Since child_open_exim() is
812 documented for local_scan(), the new function should be too.
814 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
815 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
816 results in an empty string is now treated as unset.
818 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
820 PH/31 Added sender_verify_fail logging option.
822 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
823 needed by RFC 822 but not by RFC 2822 was commented out. I have now
824 tidied the source and removed it altogether.
826 PH/33 When a queue run was abandoned because the load average was too high, a
827 log line was always written; now it is written only if the queue_run log
828 selector is set. In addition, the log line for abandonment now contains
829 information about the queue run such as the pid. This is always present
830 in "start" and "stop" lines but was omitted from the "abandon" line.
832 PH/34 Omit spaces between a header name and the colon in the error message that
833 is given when verify = headers_syntax fails (if there are lots of them,
834 the message gets confusing).
836 PH/35 Change the default for dns_check_names_pattern to allow slashes within
837 names, as there are now some PTR records that contain slashes. This check
838 is only to protect against broken name servers that fall over on strange
839 characters, so the fact that it applies to all lookups doesn't matter.
841 PH/36 Now that the new test suite is complete, we can remove some of the
842 special code in Exim that was needed for the old test suite. For example,
843 sorting DNS records because real resolvers return them in an arbitrary
844 order. The new test suite's fake resolver always returns records in the
847 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
848 bounces) except when queue_only is set, to avoid logging races between
849 the different processes.
851 PH/38 Panic-die if .include specifies a non-absolute path.
853 PH/39 A tweak to the "H" retry rule from its user.
855 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
856 a label. They prevented compilation on older perls.
858 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
859 a warning to be raised on newish perls.
861 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
862 on queue. Changes to match documented behaviour of showing count of
863 messages matching specified criteria.
865 PH/40 Changed the default ident timeout from 30s to 5s.
867 PH/41 Added support for the use of login_cap features, on those BSD systems
868 that have them, for controlling the resources used by pipe deliveries.
870 PH/42 The content-scanning code uses fopen() to create files in which to put
871 message data. Previously it was not paying any attention to the mode of
872 the files. Exim runs with umask(0) because the rest of the code creates
873 files with open(), and sets the required mode explicitly. Thus, these
874 files were ending up world-writeable. This was not a big issue, because,
875 being within the spool directory, they were not world-accessible. I have
876 created a function called modefopen, which takes an additional mode
877 argument. It sets umask(777), creates the file, chmods it to the required
878 mode, then resets the umask. All the relevant calls to fopen() in the
879 content scanning code have been changed to use this function.
881 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
882 to 24 hours. This avoids potential overflow problems when processing G
883 and H retry rules. I suspect nobody ever tinkers with this value.
885 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
887 PH/45 When the plaintext authenticator is running as a client, the server's
888 challenges are checked to ensure they are valid base64 strings. By
889 default, the authentication attempt is cancelled if an invalid string is
890 received. Setting client_ignore_invalid_base64 true ignores these errors.
891 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
892 they are received. Thus, the responses can be made to depend on the
893 challenges. If an invalid string is ignored, an empty string is placed in
896 PH/46 Messages that are created by the autoreply transport now contains a
897 References: header, in accordance with RFCs 2822 and 3834.
899 PH/47 Added authenticated_sender_force to the smtp transport.
901 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
903 PH/49 Installed latest patch from the Sieve maintainer.
905 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
906 was also set, the appendfile transport was unnecessarily scanning a
907 directory of message files (e.g. for maildir delivery) to find the count
908 of files (along with the size), even though it did not need this
909 information. It now does the scan only if it needs to find either the
910 size of the count of files.
912 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
914 PH/52 Two bugs concerned with error handling when the smtp transport is
917 (i) Exim was not creating retry information for temporary errors given
918 for individual recipients after the DATA command when the smtp transport
919 was used in LMTP mode. This meant that they could be retried too
920 frequently, and not timed out correctly.
922 (ii) Exim was setting the flag that allows error details to be returned
923 for LMTP errors on RCPT commands, but not for LMTP errors for individual
924 recipients that were returned after the DATA command.
926 PH/53 This is related to PH/52, but is more general: for any failing address,
927 when detailed error information was permitted to be returned to the
928 sender, but the error was temporary, then after the final timeout, only
929 "retry timeout exceeded" was returned. Now it returns the full error as
930 well as "retry timeout exceeded".
932 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
933 do this, and (what is worse) MTAs that accept it.
935 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
936 will now be deprecated.
938 PH/56 New os.c-cygwin from the Cygwin maintainer.
940 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
941 formats (previously only available in exim formats via -bpr, -bpru,
942 and -bpra. Now also available in native and exiqgrep formats)
944 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
945 with very large, slow to parse queues
947 JJ/08 exipick: added ! as generic prefix to negate any criteria format
949 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
951 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
952 responses to authentication challenges, though it was showing the
953 challenges; (ii) I've removed the CR characters from the debug output for
956 PH/58 Allow for the insertion of a newline as well as a space when a string
957 is turned into more than one encoded-word during RFC 2047 encoding. The
958 Sieve code now uses this.
960 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
961 data_4xx, lost_connection, tls_required.
963 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
964 message was being sent as an SMTP response.
966 PH/61 Add -l and -k options to exicyclog.
968 PH/62 When verifying, if an address was redirected to one new address, so that
969 verification continued, and the new address failed or deferred after
970 having set something in $address_data, the value of $address_data was not
971 passed back to the ACL. This was different to the case when no
972 redirection occurred. The value is now passed back in both cases.
974 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
975 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
976 use login_cap.h, so on its own it isn't the distinguishing feature. The
977 new name refers directly to the setclassresources() function.
979 PH/65 Added configuration files for NetBSD3.
981 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
983 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
984 is preferred over IPv4.
986 PH/68 The bounce_return_message and bounce_return_body options were not being
987 honoured for bounces generated during the reception of non-SMTP messages.
988 In particular, this applied to messages rejected by the ACL. This bug has
989 been fixed. However, if bounce_return_message is true and bounce_return_
990 body is false, the headers that are returned for a non-SMTP message
991 include only those that have been read before the error was detected.
992 (In the case of an ACL rejection, they have all been read.)
994 PH/69 The HTML version of the specification is now built in a directory called
995 spec_html instead of spec.html, because the latter looks like a path with
996 a MIME-type, and this confuses some software.
998 PH/70 Catch two compiler warnings in sieve.c.
1000 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1001 function verify_get_ident() calls ip_connect() to connect a socket, but
1002 if the "connect()" function timed out, ip_connect() used to close the
1003 socket. However, verify_get_ident() also closes the socket later, and in
1004 between Exim writes to the log, which may get opened at this point. When
1005 the socket was closed in ip_connect(), the log could get the same file
1006 descriptor number as the socket. This naturally causes chaos. The fix is
1007 not to close the socket in ip_connect(); the socket should be closed by
1008 the function that creates it. There was only one place in the code where
1009 this was missing, in the iplookup router, which I don't think anybody now
1010 uses, but I've fixed it anyway.
1012 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1013 well as to direct DNS lookups. Otherwise the handling of names in host
1014 lists is inconsistent and therefore confusing.
1020 PH/01 Two changes to the default runtime configuration:
1022 (1) Move the checks for relay_from_hosts and authenticated clients from
1023 after to before the (commented out) DNS black list checks.
1025 (2) Add control=submission to the relay_from_hosts and authenticated
1026 clients checks, on the grounds that messages accepted by these
1027 statements are most likely to be submissions.
1029 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1031 (1) Generate an error if the third argument for the ${prvs expansion is
1034 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1037 (3) Reset the variables that are obtained from the first argument of
1038 ${prvscheck and used in the second argument before leaving the code,
1039 because their memory is reclaimed, so using them afterwards may do
1042 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1043 one (it's much easier than Tom thought :-).
1045 (5) Because of (4), we can now allow for the use of $prvscheck_result
1046 inside the third argument.
1048 PH/03 For some reason, the default setting of PATH when running a command from
1049 a pipe transport was just "/usr/bin". I have changed it to
1052 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1053 anything to be listed in the output from -bV.
1055 PH/05 When a filter generated an autoreply, the entire To: header line was
1056 quoted in the delivery log line, like this:
1058 => >A.N.Other <ano@some.domain> <original@ddress> ...
1060 This has been changed so that it extracts the operative address. There
1061 may be more than one such address. If so, they are comma-separated, like
1064 => >ano@some.domain,ona@other.domain <original@ddress> ...
1066 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1067 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1068 not being looked up in the rDNS to get a host name, Exim was showing the
1069 IP address twice in Received: lines, even though the IP addresses were
1070 identical. For example:
1072 Received: from [1.2.3.4] (helo=[1.2.3.4])
1074 However, if the real host name was known, it was omitting the HELO data
1075 if it matched the actual IP address. This has been tidied up so that it
1076 doesn't show the same IP address twice.
1078 PH/07 When both +timestamp and +memory debugging was on, the value given by
1079 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1080 called by the debug printing, thereby overwriting the timestamp buffer.
1081 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1083 PH/08 When the original message was included in an autoreply transport, it
1084 always said "this is a copy of the message, including all the headers",
1085 even if body_only or headers_only was set. It now gives an appropriate
1088 PH/09 Applied a patch from the Sieve maintainer which:
1090 o fixes some comments
1091 o adds the (disabled) notify extension core
1092 o adds some debug output for the result of if/elsif tests
1093 o points to the current vacation draft in the documentation
1094 and documents the missing references header update
1098 o fixes a bug in processing the envelope test (when testing
1099 multiple envelope elements, the last element determinted the
1102 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1103 Electronic Mail") by including:
1105 Auto-submitted: auto-generated
1107 in the messages that it generates (bounce messages and others, such as
1108 warnings). In the case of bounce messages for non-SMTP mesages, there was
1109 also a typo: it was using "Auto_submitted" (underscore instead of
1110 hyphen). Since every message generated by Exim is necessarily in response
1111 to another message, thes have all been changed to:
1113 Auto-Submitted: auto-replied
1115 in accordance with these statements in the RFC:
1117 The auto-replied keyword:
1119 - SHOULD be used on messages sent in direct response to another
1120 message by an automatic process,
1122 - MUST NOT be used on manually-generated messages,
1124 - MAY be used on Delivery Status Notifications (DSNs) and Message
1125 Disposition Notifications (MDNs),
1127 - MUST NOT be used on messages generated by automatic or periodic
1128 processes, except for messages which are automatic responses to
1131 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1132 to the default Received: header definition.
1134 PH/12 Added log selector acl_warn_skipped (default on).
1136 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1137 variables because (a) they are in the wrong storage pool and (b) even if
1138 they were copied, it wouldn't work properly because of the caching.
1140 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1141 checking when decoding. Apparently there are clients that generate
1142 overlong encoded strings. Why am I not surprised?
1144 PH/15 If the first argument of "${if match_address" was not empty, but did not
1145 contain an "@" character, Exim crashed. Now it writes a panic log message
1146 and treats the condition as false.
1148 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1150 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1151 extension "envelope-auth". The code is finished and in agreement with
1152 other implementations, but there is no documentation so far and in fact,
1153 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1154 not changing the active code.
1156 Print executed "if" and "elsif" statements when debugging is used. This
1157 helps a great deal to understand what a filter does.
1159 Document more things not specified clearly in RFC3028. I had all this
1160 sorted out, when out of a sudden new issues came to my mind. Oops."
1162 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1165 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1166 canonical form (as documented). However, after a host name lookup from
1167 the IP address, check_host() was doing a simple string comparison with
1168 addresses acquired from the DNS when checking that the found name did
1169 have the original IP as one of its addresses. Since any found IPv6
1170 addresses are likely to be in abbreviated form, the comparison could
1171 fail. Luckily, there already exists a function for doing the comparison
1172 by converting both addresses to binary, so now that is used instead of
1173 the text comparison.
1175 PH/20 There was another similar case to PH/19, when a complete host name was
1176 given in a host list; looking up its IP address could give an abbreviated
1177 form, whereas the current host's name might or might not be abbreviated.
1178 The same fix has been applied.
1184 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1185 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1188 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1189 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1191 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1193 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1194 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1195 Also, for "Auto-Submitted", treat anything other than "no" as
1196 non-personal, in accordance with RFC 3834. (Previously it treated
1197 anything starting "auto-" as non-personal.)
1199 TF/01 The control=submission/name=... option had a problem with syntax
1200 errors if the name included a slash character. The /name= option
1201 now slurps the rest of the string, so it can include any characters
1202 but it must come last in the list of options (after /sender_retain
1205 PH/05 Some modifications to the interface to the fake nameserver for the new
1213 TK/01 Added the "success_on_redirect" address verification option. See
1214 NewStuff for rationale and an example.
1216 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1218 PH/02 Patch to exigrep to allow it to work on syslog lines.
1220 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1221 fread() to skip over the body file's header line, because in Cygwin the
1222 header line is locked and is inaccessible.
1224 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1225 co-exist for some time) to make it clear that it is the Exim ID that is
1226 referenced, not the Message-ID: header line.
1228 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1229 string_format() function, because snprintf() does not exist on all
1232 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1233 ${stat: expansion item.
1235 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1236 protocol synchronization error", to keep the pedants happy.
1238 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1239 well as for IRIX systems, when gcc is being used. See the host.c source
1242 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1244 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1247 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1248 transport and to the smtp transport in LMTP mode.
1250 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1252 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1253 run a filter in a subprocess. This could lead to confusion in subsequent
1254 lookups in the parent process. There should also be a search_tidyup() at
1255 the end of the subprocess.
1257 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1258 only if the host matched helo_try_verify_hosts, which caused the
1259 verification to occur when the EHLO/HELO command was issued. The ACL just
1260 tested the remembered result. Now, if a previous verification attempt has
1261 not happened, "verify = helo" does it there and then.
1263 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1265 TK/03 Fix log output including CR from clamd.
1267 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1268 exist provoked a memory error which could cause a segfault.
1270 PH/15 Installed PCRE 6.2
1272 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1274 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1275 of the problem. Specifically, suggested +O2 rather than +O1 for the
1278 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1280 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1281 record, the port was not correctly logged when the outgoing_port log
1282 selector was set (it logged the transort's default port).
1284 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1285 fallback_hosts, and "hosts" in the smtp transport.
1287 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1288 host errors such as "Connection refused".
1290 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1291 authentication with radiusclient 0.4.9:
1293 - Error returned from rc_read_config was caught wrongly
1294 - Username/password not passed on to radius server due to wrong length.
1296 The presumption is that some radiusclient API changes for 4.51/PH/17
1297 were not taken care of correctly. The code is still untested by me (my
1298 Linux distribution still has 0.3.2 of radiusclient), but it was
1299 contributed by a Radius user.
1301 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1302 expanding the "port" option of the smtp transport.
1304 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1305 while reading a MIME header. Thanks to Tom Hughes for a patch.
1307 PH/24 Include config.h inside local_scan.h so that configuration settings are
1310 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1311 that in an ACL for RCPT (for example), you can examine exactly what was
1314 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1315 commands, but it was not correctly comparing the address with the actual
1316 client host address. Thus, it would show the EHLO address in Received:
1317 header lines when this was not necessary.
1319 PH/27 Added the % operator to ${eval:}.
1321 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1322 it should be ignoring failures (because with -C, for example, it has lost
1323 privilege). It wasn't ignoring creation failures other than "already
1326 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1329 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1330 because an input line was too long, either on its own, or by virtue of
1331 too many continuations, the temporary file was not being removed, and the
1332 return code was incorrect.
1334 PH/31 Missing "BOOL" in function definition in filtertest.c.
1336 PH/32 Applied Sieve patches from the maintainer.
1338 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1340 PH/33 Added "verify = not_blind".
1342 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1343 Local/Makefile (with some defaults set). These are used in built scripts
1344 such as exicyclog, but they have never been used in the exim_install
1345 script (though there are many overriding facilities there). I have
1346 arranged that the exim_install script now takes note of these two
1349 PH/35 Installed configuration files for Dragonfly.
1351 PH/36 When a locally submitted message by a trusted user did not contain a
1352 From: header, and the sender address was obtained from -f or from an SMTP
1353 MAIL command, and the trusted user did not use -F to supply a sender
1354 name, $originator_name was incorrectly used when constructing a From:
1355 header. Furthermore, $originator_name was used for submission mode
1356 messages from external hosts without From: headers in a similar way,
1357 which is clearly wrong.
1359 PH/37 Added control=suppress_local_fixups.
1361 PH/38 When log_selector = +received_sender was set, and the addition of the
1362 sender made the log line's construction buffer exactly full, or one byte
1363 less than full, an overflow happened when the terminating "\n" was
1366 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1367 when the result of a list match is failure because a DNS lookup failed.
1369 PH/40 RM_COMMAND is now used in the building process.
1371 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1372 the "build-* directories that it finds.
1374 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1375 address in a domain literal was a prefix of an interface address.
1377 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1378 when verifying a sender address, unless rewrite_headers is false.
1380 PH/44 Wrote a long comment about why errors_to addresses are verified as
1381 recipients, not senders.
1383 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1384 the ratelimit ACL was added.
1386 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1388 PH/46 Added extra information about PostgreSQL errors to the error string.
1390 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1391 avoiding the need to install special zones in a real server. This is
1392 backwards compatible; if it can't find the fake resolver, it drops back.
1393 Thus, both old and new test suites can be run.
1395 TF/02 Added util/ratelimit.pl
1397 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1398 clock is set back in time.
1400 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1401 Candler <B.Candler@pobox.com>.
1403 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1404 OK for addresses that are the result of redirections.
1406 PH/48 A number of further additions for the benefit of the new test suite,
1407 including a fake gethostbyname() that interfaces to the fake DNS resolver
1410 TF/06 The fix for widen_domains has also been applied to qualify_single and
1411 search_parents which are the other dnslookup options that can cause
1414 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1417 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1419 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1420 service" if it encountered an MX record with an empty target hostname.
1421 The message is now "an MX or SRV record indicated no SMTP service".
1423 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1424 if the DNS of the sending site is misconfigured. This is quite a
1425 common situation. This change restores the behaviour of treating a
1426 helo verification defer as a failure.
1428 PH/51 If self=fail was set on a router, the bounce message did not include the
1429 actual error message.
1435 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1437 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1438 command itself ended in error, the underlying message about the transport
1439 filter timeout was being overwritten with the pipe command error. Now the
1440 underlying error message should be appended to the second error message.
1442 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1444 PH/02 Reduce the amount of output that "make" produces by default. Full output
1445 can still be requested.
1447 PH/03 The warning log line about a condition test deferring for a "warn" verb
1448 was being output only once per connection, rather than after each
1449 occurrence (because it was using the same function as for successful
1450 "warn" verbs). This seems wrong, so I have changed it.
1452 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1453 it should not have, which might have caused a crash in the right
1454 circumstances, but probably never did.
1456 PH/04 Installed a modified version of Tony Finch's patch to make submission
1457 mode fix the return path as well as the Sender: header line, and to
1458 add a /name= option so that you can make the user's friendly name appear
1461 TF/03 Added the control = fakedefer ACL modifier.
1463 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1464 Mark Lowes for thorough testing.
1466 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1468 TK/03 Merged latest SRS patch from Miles Wilton.
1470 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1471 with the definition in sysexits.h (which is #included earlier).
1472 Fortunately, Exim does not actually use EX_OK. The code used to try to
1473 preserve the sysexits.h value, by assumimg that macro definitions were
1474 scanned for macro replacements. I have been disabused of this notion,
1475 so now the code just undefines EX_OK before #including unistd.h.
1477 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1478 in the smtp transport. When a block could not be written in a single
1479 write() function, the timeout was being re-applied to each part-write.
1480 This seems wrong - if the receiver was accepting one byte at a time it
1481 would take for ever. The timeout is now adjusted when this happens. It
1482 doesn't have to be particularly precise.
1484 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1485 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1487 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1488 without a domain if the check to <postmaster@domain> fails.
1490 SC/01 Eximstats: added -xls and the ability to specify output files
1491 (patch written by Frank Heydlauf).
1493 SC/02 Eximstats: use FileHandles for outputing results.
1495 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1497 SC/04 Eximstats: fixed display of large numbers with -nvr option
1499 SC/05 Eximstats: fixed merging of reports with empty tables.
1501 SC/06 Eximstats: added the -include_original_destination flag
1503 SC/07 Eximstats: removed tabs and trailing whitespace.
1505 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1507 TK/06 MBOX spool code: Add real "From " MBOX separator line
1508 so the .eml file is really in mbox format (even though
1509 most programs do not really care). Patch from Alex Miller.
1511 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1512 The latter is generated from $received_to and is only set if the
1513 message has one envelope recipient. SA can use these headers,
1514 obviously out-of-the-box. Patch from Alex Miller.
1516 PH/08 The ${def test on a variable was returning false if the variable's
1517 value was "0", contrary to what the specification has always said!
1518 The result should be true unless the variable is empty.
1520 PH/09 The syntax error of a character other than { following "${if
1521 def:variable_name" (after optional whitespace) was not being diagnosed.
1522 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1523 accidental colon was present, for example, could give incorrect results.
1525 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1526 result is used (not including appendfile, where other changes are about
1529 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1530 This involved changing a lot of size variables from int to off_t. It
1531 should work with maildirs and everything.
1533 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1534 spamd dying while we are connected to it.
1536 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1539 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1540 function for BDB 4.3.
1542 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1544 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1545 the sources for PCRE have been re-arranged and more files are now
1548 PH/15 The code I had for printing potentially long long variables in PH/11
1549 above was not the best (it lost precision). The length of off_t variables
1550 is now inspected at build time, and an appropriate printing format (%ld
1551 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1552 to be "long long int" or "long int". This is needed for the internal
1553 formatting function string_vformat().
1555 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1556 the configuration file to be ":syslog", then the script "guesses" where
1557 the logs files are, rather than using the compiled in default. In our
1558 case the guess is not the same as the compiled default, so the script
1559 suddenly stopped working when I started to use syslog. The patch checks
1560 to see if log_file_path is "". If so, it attempts to read it from exim
1561 with no configuration file to get the compiled in version, before it
1562 falls back to the previous guessing code."
1564 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1565 implementing BATV in an Exim configuration. See NewStuff for the gory
1568 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1569 Makefile that are specific to HP-UX.
1571 PH/18 If the "use_postmaster" option was set for a recipient callout together
1572 with the "random" option, the postmaster address was used as the MAIL
1573 FROM address for the random test, but not for the subsequent recipient
1574 test. It is now used for both.
1576 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1577 patch removes a few documentation additions to RFC 3028, because the
1578 latest draft now contains them. It adds the new en;ascii-case comparator
1579 and a new error check for 8bit text in MIME parts. Comparator and
1580 require names are now matched exactly. I enabled the subaddress
1581 extension, but it is not well tested yet (read: it works for me)."
1583 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1584 rework some of the code of TK/09 above to avoid the hardwired use of
1585 "%lld" and "long long". Replaced the call to snprintf() with a call to
1588 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1589 records point to non-existent hosts", "retry timeout exceeded", and
1590 "retry time not reached for any host after a long failure period".
1592 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1593 experimental DomainKeys support:
1595 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1596 (2) On an error such as an illegally used "control", the wrong name for
1597 the control was given.
1599 These problems did NOT occur unless DomainKeys support was compiled.
1601 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1603 PH/24 Added ${if match_ip condition.
1605 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1606 databases so that it will be absolutely obvious if a crash occurs in the
1607 DB library. This is a regular occurrence (often caused by mis-matched
1610 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1611 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1612 picked up on a user's system that detects such things. There doesn't seem
1613 to be a gcc warning option for this - only an attribute that has to be
1614 put on the function's prototype. It seems that in Fedora Core 4 they have
1615 set this on a number of new functions. No doubt there will be more in due
1618 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1619 specify a transport. However, if an address that was verified by such a
1620 router was the subject of a callout, Exim crashed because it tried to
1621 read the rcpt_include_affixes from the non-existent transport. Now it
1622 just assumes that the setting of that option is false. This bug was
1623 introduced by 4.51/PH/31.
1625 PH/28 Changed -d+all to exclude +memory, because that information is very
1626 rarely of interest, but it makes the output a lot bigger. People tend to
1627 do -d+all out of habit.
1629 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1630 code in os-type was giving problems when libc.so lives in lib64, like on
1633 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1634 aren't the modern standard, and it seems that some systems' include files
1635 don't always have them. Exim was already checking for some of the newer
1636 ones like T_AAAA, and defining it itself. I've added checks for all the
1637 record types that Exim uses.
1639 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1640 not automatically generating a new one, as it is supposed to. This
1641 prevented TLS from working. If the file did exist, but contained invalid
1642 data, a new version was generated, as expected. It was only the case of a
1643 non-existent file that was broken.
1645 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1646 with a change in libdomainkeys > 0.64.
1648 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1649 from DNS. If the selector record carries the flag, it now has
1650 precedence over the domain-wide flag.
1652 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1654 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1655 the use of a port name that isn't defined in /etc/services) occurred, the
1656 message was deferred as in a normal delivery, and thus remained on the
1657 spool, instead of being failed because of the mua_wrapper setting. This
1658 is now fixed, and I tidied up some of the mua_wrapper messages at the
1661 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1662 the messages in a hash of arrays rather than using individual hashes.
1663 This is a bit cleaner and results in dramatic memory savings, albeit
1664 at a slight CPU cost.
1666 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1667 as requested by Marc Sherman.
1669 SC/10 Eximstats: added histograms for user specified patterns as requested
1672 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1674 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1675 fopen() in the content-scanning modules that did not already have it.
1681 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1682 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1684 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1686 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1688 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1689 to test Sieve filters that use "vacation".
1691 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1692 that changes the way the GnuTLS parameters are stored in the cache file.
1693 The new format can be generated externally. For backward compatibility,
1694 if the data in the cache doesn't make sense, Exim assumes it has read an
1695 old-format file, and it generates new data and writes a new file. This
1696 means that you can't go back to an older release without removing the
1699 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1700 work if there are any delivery delays because "one_time" forces the
1701 parent to be marked "delivered", so its unseen clone is never tried
1702 again. For this reason, Exim now forbids the simultaneous setting of
1705 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1706 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1707 entry if you want to know the details. The fix, however, neglected to
1708 consider the case when local delivery batching is involved. The test for
1709 "previously delivered" was not happening when checking to see if an
1710 address could be batched with a previous (undelivered) one; under
1711 certain circumstances this could lead to multiple deliveries to the same
1714 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1715 in its include files, and this causes problems building Exim.
1717 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1718 header_syntax) but Exim was just ignoring anything given after a slash.
1719 In particular, this caused confusion with an attempt to use "verify =
1720 reverse_host_lookup/defer_ok". An error is now given when options are
1721 supplied for verify items that do not have them. (Maybe reverse_host_
1722 lookup should have a defer_ok option, but that's a different point.)
1724 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1725 defined by RFC 821) to 2048, because there were problems with some AUTH
1726 commands, and RFC 1869 says the size should be increased for extended
1727 SMTP commands that take arguments.
1729 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1732 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1733 "unknown" error; now it says that the functionality isn't in the binary.
1735 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1736 an address' error message when a string expansion fails (syntax or
1737 whatever). Otherwise the password may appear in the log. Following change
1738 PH/42 below, there is no longer a chance of it appearing in a bounce
1741 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1743 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1744 of its addresses was ever tried. (Bugzilla bug #2.)
1746 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1747 the result incorrectly in the debug output. (It correctly added a newline
1748 to what was transported.)
1750 TF/01 Added $received_time.
1752 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1753 commented out examples of how to interface to a virus scanner and to
1754 SpamAssassin. Also added commented examples of av_scanner and
1755 spamd_address settings.
1757 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1758 and controls are allowed in which ACLs. There were a couple of minor
1759 errors. Some of the entries in the conditions table (which is a table of
1760 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1761 negation of where the condition IS allowed.
1763 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1765 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1766 header file does not have a version number, so I've had to invent a new
1767 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1768 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1769 radiusclient), but it was contributed by a Radius user.
1771 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1772 files or directories, for OpenSSL.
1774 PH/19 When an Exim process that is running as root has to create an Exim log
1775 file, it does so in a subprocess that runs as exim:exim so as to get the
1776 ownership right at creation (otherwise, other Exim processes might see
1777 the file with the wrong ownership). There was no test for failure of this
1778 fork() call, which would lead to the process getting stuck as it waited
1779 for a non-existent subprocess. Forks do occasionally fail when resources
1780 run out. I reviewed all the other calls to fork(); they all seem to check
1783 PH/20 When checking for unexpected SMTP input at connect time (before writing
1784 the banner), Exim was not dealing correctly with a non-positive return
1785 from the read() function. If the client had disconnected by this time,
1786 the result was a log entry for a synchronization error with an empty
1787 string after "input=" when read() returned zero. If read() returned -1
1788 (an event I could not check), uninitialized data bytes were printed.
1789 There were reports of junk text (parts of files, etc) appearing after
1792 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1794 PH/22 Added support for macro redefinition, and (re)definition in between
1795 driver and ACL definitions.
1797 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1798 forgetting to use the resulting value; it was using the unexpanded value.
1800 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1801 hadn't been configured. The fix is from Juergen Kreileder, who
1802 understands it better than I do:
1804 "Here's what I see happening with three configured cyrus_sasl
1805 authenticators configured (plain, login, cram-md5):
1807 On startup auth_cyrus_sasl_init() gets called for each of these.
1808 This means three calls to sasl_listmech() without a specified mech_list.
1809 => SASL tests which mechs of all available mechs actually work
1810 => three warnings about OTP not working
1811 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1813 With the patch, sasl_listmech() also gets called three times. But now
1814 SASL's mech_list option is set to the server_mech specified in the the
1815 authenticator. Or in other words, the answer from sasl_listmech()
1816 gets limited to just the mech you're testing for (which is different
1818 => the return list contains just 'plain' or 'login', 'cram-md5' or
1819 nothing depending on the value of ob->server_mech.
1821 I've just tested the patch: Authentication still works fine,
1822 unavailable mechs specified in the exim configuration are still
1823 caught, and the auth.log warnings about OTP are gone."
1825 PH/25 When debugging is enabled, the contents of the command line are added
1826 to the debugging output, even when log_selector=+arguments is not
1829 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1830 answer is "GNU", and only if the return is "GNU/something" is the answer
1833 PH/27 $acl_verify_message is now set immediately after the failure of a
1834 verification in an ACL, and so is available in subsequent modifiers. In
1835 particular, the message can be preserved by coding like this:
1837 warn !verify = sender
1838 set acl_m0 = $acl_verify_message
1840 Previously, $acl_verify_message was set only while expanding "message"
1841 and "log_message" when a very denied access.
1843 PH/28 Modified OS/os.c-Linux with
1845 -#ifndef OS_LOAD_AVERAGE
1846 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1848 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1849 nomenclature these days.)
1851 PH/29 Installed patch from the Sieve maintainer that adds the options
1852 sieve_useraddress and sieve_subaddress to the redirect router.
1854 PH/30 In these circumstances:
1855 . Two addresses routed to the same list of hosts;
1856 . First host does not offer TLS;
1857 . First host accepts first address;
1858 . First host gives temporary error to second address;
1859 . Second host offers TLS and a TLS session is established;
1860 . Second host accepts second address.
1861 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1862 and peerdn, if requested) that were in fact used only for the second
1865 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1866 attention to any local part prefix or suffix that was matched by the
1867 router that accepted the address. It now behaves in the same way as it
1868 does for delivery: the affixes are removed from the local part unless
1869 rcpt_include_affixes is set on the transport.
1871 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1872 timeout during the DATA phase of an incoming message.
1874 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1875 applied a patch sanctioned by the Sieve maintainer.
1877 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1878 the uid or gid is negative. A case of a negative gid caused this to be
1879 noticed. The fix allows for either to be negative.
1881 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1882 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1885 PH/36 The change PH/12 above was broken. Fixed it.
1887 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1888 the grounds that routing the same address twice would always produce the
1889 same answer. This might have been true once, but it is certainly no
1890 longer true now. Routing a child address may depend on the previous
1891 routing that produced that child. Some complicated redirection strategies
1892 went wrong when messages had multiple recipients, and made Exim's
1893 behaviour dependent on the order in which the addresses were given.
1895 I have moved the duplicate checking until after the routing is complete.
1896 Exim scans the addresses that are assigned to local and remote
1897 transports, and removes any duplicates. This means that more work will be
1898 done, as duplicates will always all be routed, but duplicates are
1899 presumably rare, so I don't expect this is of any significance.
1901 For deliveries to pipes, files, and autoreplies, the duplicate checking
1902 still happens during the routing process, since they are not going to be
1905 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1906 It corrects a timeout issue with spamd. This is Ian's comment: "The
1907 background is that sometimes spamd either never reads data from a
1908 connection it has accepted, or it never writes response data. The exiscan
1909 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1910 blindly assumes that writes won't block so it may never time out."
1912 PH/39 Allow G after quota size as well as K and M.
1914 PH/40 The value set for $authenticated_id in an authenticator may not contain
1915 binary zeroes or newlines because the value is written to log lines and
1916 to spool files. There was no check on this. Now the value is run through
1917 the string_printing() function so that such characters are converted to
1918 printable escape sequences.
1920 PH/41 $message_linecount is a new variable that contains the total number of
1921 lines in the message. Compare $body_linecount, which is the count for the
1924 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1925 bounce and delay warning messages, except in certain special cases, which
1928 (a) An SMTP error message from a remote host;
1929 (b) A message specified in a :fail: redirection;
1930 (c) A message specified in a "fail" command in a system filter;
1931 (d) A message specified in a FAIL return from the queryprogram router;
1932 (e) A message specified by the cannot_route_message router option.
1934 In these cases only, Exim does include the error details in bounce and
1935 warning messages. There are also a few cases where bland messages such
1936 as "unrouteable address" or "local delivery error" are given.
1938 PH/43 $value is now also set for the "else" part of a ${run expansion.
1940 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1941 being worked on, but at least Exim now implements the latest version to
1944 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1945 process to complete was treated as a delivery failure, a timeout while
1946 writing the message to the pipe was logged, but erroneously treated as a
1947 successful delivery. Such timeouts include transport filter timeouts. For
1948 consistency with the overall process timeout, these timeouts are now
1949 treated as errors, giving rise to delivery failures by default. However,
1950 there is now a new Boolean option for the pipe transport called
1951 timeout_defer, which, if set TRUE, converts the failures into defers for
1952 both kinds of timeout. A transport filter timeout is now identified in
1955 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1956 systems where "make" and "gmake" are different, calling "gmake" at top
1957 level broke things. I've arranged for the value of $(MAKE) to be passed
1958 from the Makefile to this script so that it can call the same version of
1962 A note about Exim versions 4.44 and 4.50
1963 ----------------------------------------
1965 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1966 changes of various kinds. As a consequence, a big documentation update was
1967 needed. This delayed the release for rather longer than seemed good, especially
1968 in the light of a couple of (minor) security issues. Therefore, the changes
1969 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1970 release. So 4.44 and 4.50 are in effect two different branches that both start
1973 I have left the 4.50 change log unchanged; it contains all the changes since
1974 4.43. The change log for 4.44 is below; many of its items are identical to
1975 those for 4.50. This seems to be the most sensible way to preserve the
1976 historical information.
1982 1. Minor wording change to the doc/README.SIEVE file.
1984 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1985 computation of the current number of files was incorrect.
1987 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1988 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1991 4. Give more explanation in the error message when the command for a transport
1992 filter fails to execute.
1994 5. There are several places where Exim runs a non-Exim command in a
1995 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1996 was being done only for the command run by the queryprogram router. It is
1997 now done for all such subprocesses. The other cases are: ${run, transport
1998 filters, and the commands run by the lmtp and pipe transports.
2000 6. Added CONFIGURE_GROUP build-time option.
2002 7. Some older OS have a limit of 256 on the maximum number of file
2003 descriptors. Exim was using setrlimit() to set 1000 as a large value
2004 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2005 systems. I've change it so that if it can't get 1000, it tries for 256.
2007 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2008 was an oversight, and furthermore, ever since the addition of extra
2009 controls (e.g. 4.43/32), the checks on when to allow different forms of
2010 "control" were broken. There should now be diagnostics for all cases when a
2011 control that does not make sense is encountered.
2013 9. Added the /retain_sender option to "control=submission".
2015 10. $recipients is now available in the predata ACL (oversight).
2017 11. Tidy the search cache before the fork to do a delivery from a message
2018 received from the command line. Otherwise the child will trigger a lookup
2019 failure and thereby defer the delivery if it tries to use (for example) a
2020 cached ldap connection that the parent has called unbind on.
2022 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2023 of $address_data from the recipient verification was clobbered by the
2024 sender verification.
2026 13. The value of address_data from a sender verification is now available in
2027 $sender_address_data in subsequent conditions in the ACL statement.
2029 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2031 15. Added a new option "connect=<time>" to callout options, to set a different
2034 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2035 was its contents. (It was OK if the option was not defined at all.)
2037 17. A "Completed" log line is now written for messages that are removed from
2038 the spool by the -Mrm option.
2040 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2041 information about exactly what failed.
2043 19. Added -dd to debug only the daemon process.
2045 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2046 handles timeouts, both on the server side and network timeouts. Renamed the
2047 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2049 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2050 It is now set to "smtps".
2052 22. $host_address is now set to the target address during the checking of
2053 ignore_target_hosts.
2055 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2056 being passed; this would have caused $sender_host_name to have been used if
2057 matching the list had actually called for a host name (not very likely,
2058 since this list is usually IP addresses). A host name is now passed as
2061 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2062 code that specifies a non-restarting handler (typically sigaction() in
2063 modern systems) in an attempt to fix a rare and obscure crash bug.
2065 25. Narrowed the window for a race in the daemon that could cause it to ignore
2066 SIGCHLD signals. This is not a major problem, because they are used only to
2067 wake it up if nothing else does.
2069 26. A malformed maildirsize file could cause Exim to calculate negative values
2070 for the mailbox size or file count. Odd effects could occur as a result.
2071 The maildirsize information is now recalculated if the size or filecount
2074 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2075 support for a long time. Removed HAVE_SYS_VFS_H.
2077 28. Installed the latest version of exipick from John Jetmore.
2079 29. In an address list, if the pattern was not a regular expression, an empty
2080 subject address (from a bounce message) matched only if the pattern was an
2081 empty string. Non-empty patterns were not even tested. This was the wrong
2082 because it is perfectly reasonable to use an empty address as part of a
2083 database query. An empty address is now tested by patterns that are
2084 lookups. However, all the other forms of pattern expect the subject to
2085 contain a local part and a domain, and therefore, for them, an empty
2086 address still always fails if the pattern is not itself empty.
2088 30. Exim went into a mad DNS loop when attempting to do a callout where the
2089 host was specified on an smtp transport, and looking it up yielded more
2090 than one IP address.
2092 31. Re-factored the code for checking spool and log partition space into a
2093 function that finds that data and another that does the check. The former
2094 is then used to implement four new variables: $spool_space, $log_space,
2095 $spool_inodes, and $log_inodes.
2097 32. The RFC2047 encoding function was originally intended for short strings
2098 such as real names; it was not keeping to the 75-character limit for
2099 encoded words that the RFC imposes. It now respects the limit, and
2100 generates multiple encoded words if necessary. To be on the safe side, I
2101 have increased the buffer size for the ${rfc2047: expansion operator from
2104 33. It is now permitted to omit both strings after an "if" condition; if the
2105 condition is true, the result is "true". As before, when the second string
2106 is omitted, a false condition yields an empty string. This makes it less
2107 cumbersome to write custom ACL and router conditions.
2109 34. Failure to deliver a bounce message always caused it to be frozen, even if
2110 there was an errors_to setting on the router. The errors_to setting is now
2113 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2114 canonical form (fully expanded) before being placed in
2115 $sender_host_address.
2117 36. The table in the code that translates DNS record types into text (T_A to
2118 "A" for instance) was missing entries for NS and CNAME. It is just possible
2119 that this could have caused confusion if both these types were looked up
2120 for the same domain, because the text type is used as part of Exim's
2121 per-process caching. But the chance of anyone hitting this buglet seems
2124 37. The dnsdb lookup has been extended in a number of ways.
2126 (1) There is a new type, "zns", which walks up the domain tree until it
2127 finds some nameserver records. It should be used with care.
2129 (2) There is a new type, "mxh", which is like "mx" except that it returns
2130 just the host names, not the priorities.
2132 (3) It is now possible to give a list of domains (or IP addresses) to be
2133 looked up. The behaviour when one of the lookups defers can be
2134 controlled by a keyword.
2136 (4) It is now possible to specify the separator character for use when
2137 multiple records are returned.
2139 38. The dnslists ACL condition has been extended: it is now possible to supply
2140 a list of IP addresses and/or domains to be looked up in a particular DNS
2143 39. Added log_selector=+queue_time_overall.
2145 40. When running the queue in the test harness, wait just a tad after forking a
2146 delivery process, to get repeatability of debugging output.
2148 41. Include certificate and key file names in error message when GnuTLS fails
2149 to set them up, because the GnuTLS error message doesn't include the name
2150 of the failing file when there is a problem reading it.
2152 42. Allow both -bf and -bF in the same test run.
2154 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2156 44. The "Exiscan patch" is now merged into the mainline Exim source.
2158 45. Sometimes the final signoff response after QUIT could fail to get
2159 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2160 < 0 before doing a fflush(). This bug looks as though it goes back to the
2161 introduction of TLS in release 3.20, but "sometimes" must have been rare
2162 because the tests only now provoked it.
2164 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2165 (this can affect the format of dates).
2167 47. exim_tidydb, when checking for the continued existence of a message for
2168 which it has found a message-specific retry record, was not finding
2169 messages that were in split spool directories. Consequently, it was
2170 deleting retry records that should have stayed in existence.
2172 48. Steve fixed some bugs in eximstats.
2174 49. The SPA authentication driver was not abandoning authentication and moving
2175 on to the next authenticator when an expansion was forced to fail,
2176 contradicting the general specification for all authenticators. Instead it
2177 was generating a temporary error. It now behaves as specified.
2179 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2180 (the order specifies the preference for clients). The order is now AES256,
2181 AES128, 3DES, ARCFOUR128.
2183 51. Small patch to Sieve code - explicitly set From: when generating an
2186 52. Exim crashed if a remote delivery caused a very long error message to be
2187 recorded - for instance if somebody sent an entire SpamAssassin report back
2188 as a large number of 550 error lines. This bug was coincidentally fixed by
2189 increasing the size of one of Exim's internal buffers (big_buffer) that
2190 happened as part of the Exiscan merge. However, to be on the safe side, I
2191 have made the code more robust (and fixed the comments that describe what
2194 53. Now that there can be additional text after "Completed" in log lines (if
2195 the queue_time_overall log selector is set), a one-byte patch to exigrep
2196 was needed to allow it to recognize "Completed" as not the last thing in
2199 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2200 patch that reportedly fixes this has been added. I am not expert enough to
2201 create a test for it. This is what the patch creator wrote:
2203 "I found a little strange behaviour of ldap code when working with
2204 Windows 2003 AD Domain, where users was placed in more than one
2205 Organization Units. When I tried to give exim partial DN, the exit code
2206 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2207 But simultaneously result of request was absolutely normal ldap result,
2208 so I produce this patch..."
2210 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2211 REFERENCE, so I have modified the code to exclude the patch when that macro
2214 55. Some experimental protocols are using DNS PTR records for new purposes. The
2215 keys for these records are domain names, not reversed IP addresses. The
2216 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2217 leaves it alone. Component reversal etc. now happens only for IP addresses.
2220 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2222 57. Double the size of the debug message buffer (to 2048) so that more of very
2223 long debug lines gets shown.
2225 58. The exicyclog utility now does better if the number of log files to keep
2226 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2228 59. Two changes related to the smtp_active_hostname option:
2230 (1) $smtp_active_hostname is now available as a variable.
2231 (2) The default for smtp_banner uses $smtp_active_hostname instead
2232 of $primary_hostname.
2234 60. The host_aton() function is supposed to be passed a string that is known
2235 to be a valid IP address. However, in the case of IPv6 addresses, it was
2236 not checking this. This is a hostage to fortune. Exim now panics and dies
2237 if the condition is not met. A case was found where this could be provoked
2238 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2239 components; fortuitously, this particular loophole had already been fixed
2240 by change 4.50/55 above.
2242 If there are any other similar loopholes, the new check in host_aton()
2243 itself should stop them being exploited. The report I received stated that
2244 data on the command line could provoke the exploit when Exim was running as
2245 exim, but did not say which command line option was involved. All I could
2246 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2247 running as the user.
2250 61. There was a buffer overflow vulnerability in the SPA authentication code
2251 (which came originally from the Samba project). I have added a test to the
2252 spa_base64_to_bits() function which I hope fixes it.
2255 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2256 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2258 63. The daemon start-up calls getloadavg() while still root for those OS that
2259 need the first call to be done as root, but it missed one case: when
2260 deliver_queue_load_max is set with deliver_drop_privilege. This is
2261 necessary for the benefit of the queue runner, because there is no re-exec
2262 when deliver_drop_privilege is set.
2264 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2265 This has been fixed.
2267 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2268 was in use, was not putting the data itself into the right store pool;
2269 consequently, it could be overwritten for a subsequent message in the same
2270 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2273 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2275 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2276 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2277 the function were treating the return as a boolean value, which happened to
2278 work because 0=false and not-0=true, but is not correct code.
2280 68. The host_aton() function was not handling scoped IPv6 addresses (those
2281 with, for example, "%eth0" on the end) correctly.
2283 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2284 negated items (that is, ~something) in unsigned ints. Some compilers
2285 apparently mutter when there is no cast.
2287 70. If an address verification called from an ACL failed, and did not produce a
2288 user-specific message (i.e. there was only a "system" message), nothing was
2289 put in $acl_verify_message. In this situation, it now puts the system
2292 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2293 session; change 4.31/43 added the unwanted input to the log line - except
2294 that it did not do this in the start of session case. It now does.
2296 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2297 This is wrong and can cause the other end to generate a synchronization
2298 error if it is another Exim or anything else that does the synchronization
2299 check. A QUIT command is no longer sent after a timeout.
2301 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2302 during host lookups.
2304 74. The defer_ok option of callout verification was not working if it was used
2305 when verifying addresses in header lines, that is, for this case:
2307 verify = header_sender/callout=defer_ok
2309 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2310 those file descriptors could be used for SMTP connections. If anything
2311 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2312 could be sent to the SMTP client, causing chaos. The daemon now opens
2313 stdin, stdout, and stderr to /dev/null when it puts itself into the
2316 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2317 log by default. The user can override this with suitable Perl magic.
2319 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2320 the log message when discard triggers, was not working for the DATA ACL or
2321 for the non-SMTP ACL.
2323 78. Error message wording change in sieve.c.
2325 79. If smtp_accept_max_per_host was set, the number of connections could be
2326 restricted to fewer than expected, because the daemon was trying to set up
2327 a new connection before checking whether the processes handling previous
2328 connections had finished. The check for completed processes is now done
2329 earlier. On busy systems, this bug wouldn't be noticed because something
2330 else would have woken the daemon, and it would have reaped the completed
2333 80. If a message was submitted locally by a user whose login name contained one
2334 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2335 It caused a spool format error. I have fixed the spool reading code. A
2336 related problem was that the "from" clause in the Received: line became
2337 illegal because of the space(s). It is now covered by ${quote_local_part.
2339 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2342 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2344 83. Patch from Sieve maintainer to fix unterminated string problem in
2345 "vacation" handling.
2347 84. Some minor changes to the Linux configuration files to help with other
2348 OS variants using glibc.
2350 85. One more patch for Sieve to update vacation handling to latest spec.
2353 ----------------------------------------------------
2354 See the note above about the 4.44 and 4.50 releases.
2355 ----------------------------------------------------
2361 1. Change 4.43/35 introduced a bug that caused file counts to be
2362 incorrectly computed when quota_filecount was set in an appendfile
2365 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2366 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2369 3. Give more explanation in the error message when the command for a transport
2370 filter fails to execute.
2372 4. There are several places where Exim runs a non-Exim command in a
2373 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2374 was being done only for the command run by the queryprogram router. It is
2375 now done for all such subprocesses. The other cases are: ${run, transport
2376 filters, and the commands run by the lmtp and pipe transports.
2378 5. Some older OS have a limit of 256 on the maximum number of file
2379 descriptors. Exim was using setrlimit() to set 1000 as a large value
2380 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2381 systems. I've change it so that if it can't get 1000, it tries for 256.
2383 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2384 was an oversight, and furthermore, ever since the addition of extra
2385 controls (e.g. 4.43/32), the checks on when to allow different forms of
2386 "control" were broken. There should now be diagnostics for all cases when a
2387 control that does not make sense is encountered.
2389 7. $recipients is now available in the predata ACL (oversight).
2391 8. Tidy the search cache before the fork to do a delivery from a message
2392 received from the command line. Otherwise the child will trigger a lookup
2393 failure and thereby defer the delivery if it tries to use (for example) a
2394 cached ldap connection that the parent has called unbind on.
2396 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2397 of $address_data from the recipient verification was clobbered by the
2398 sender verification.
2400 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2401 was its contents. (It was OK if the option was not defined at all.)
2403 11. A "Completed" log line is now written for messages that are removed from
2404 the spool by the -Mrm option.
2406 12. $host_address is now set to the target address during the checking of
2407 ignore_target_hosts.
2409 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2410 being passed; this would have caused $sender_host_name to have been used if
2411 matching the list had actually called for a host name (not very likely,
2412 since this list is usually IP addresses). A host name is now passed as
2415 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2416 code that specifies a non-restarting handler (typically sigaction() in
2417 modern systems) in an attempt to fix a rare and obscure crash bug.
2419 15. Narrowed the window for a race in the daemon that could cause it to ignore
2420 SIGCHLD signals. This is not a major problem, because they are used only to
2421 wake it up if nothing else does.
2423 16. A malformed maildirsize file could cause Exim to calculate negative values
2424 for the mailbox size or file count. Odd effects could occur as a result.
2425 The maildirsize information is now recalculated if the size or filecount
2428 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2429 support for a long time. Removed HAVE_SYS_VFS_H.
2431 18. Updated exipick to current release from John Jetmore.
2433 19. Allow an empty sender to be matched against a lookup in an address list.
2434 Previously the only cases considered were a regular expression, or an
2437 20. Exim went into a mad DNS lookup loop when doing a callout where the
2438 host was specified on the transport, if the DNS lookup yielded more than
2441 21. The RFC2047 encoding function was originally intended for short strings
2442 such as real names; it was not keeping to the 75-character limit for
2443 encoded words that the RFC imposes. It now respects the limit, and
2444 generates multiple encoded words if necessary. To be on the safe side, I
2445 have increased the buffer size for the ${rfc2047: expansion operator from
2448 22. Failure to deliver a bounce message always caused it to be frozen, even if
2449 there was an errors_to setting on the router. The errors_to setting is now
2452 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2453 canonical form (fully expanded) before being placed in
2454 $sender_host_address.
2456 24. Updated eximstats to version 1.33
2458 25. Include certificate and key file names in error message when GnuTLS fails
2459 to set them up, because the GnuTLS error message doesn't include the name
2460 of the failing file when there is a problem reading it.
2462 26. Expand error message when OpenSSL has problems setting up cert/key files.
2465 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2466 (this can affect the format of dates).
2468 28. exim_tidydb, when checking for the continued existence of a message for
2469 which it has found a message-specific retry record, was not finding
2470 messages that were in split spool directories. Consequently, it was
2471 deleting retry records that should have stayed in existence.
2473 29. eximstats updated to version 1.35
2474 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2475 1.35 - bugfix such that pie charts by volume are generated correctly
2477 30. The SPA authentication driver was not abandoning authentication and moving
2478 on to the next authenticator when an expansion was forced to fail,
2479 contradicting the general specification for all authenticators. Instead it
2480 was generating a temporary error. It now behaves as specified.
2482 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2483 (the order specifies the preference for clients). The order is now AES256,
2484 AES128, 3DES, ARCFOUR128.
2486 31. Small patch to Sieve code - explicitly set From: when generating an
2489 32. Exim crashed if a remote delivery caused a very long error message to be
2490 recorded - for instance if somebody sent an entire SpamAssassin report back
2491 as a large number of 550 error lines. This bug was coincidentally fixed by
2492 increasing the size of one of Exim's internal buffers (big_buffer) that
2493 happened as part of the Exiscan merge. However, to be on the safe side, I
2494 have made the code more robust (and fixed the comments that describe what
2497 33. Some experimental protocols are using DNS PTR records for new purposes. The
2498 keys for these records are domain names, not reversed IP addresses. The
2499 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2500 leaves it alone. Component reversal etc. now happens only for IP addresses.
2503 34. The host_aton() function is supposed to be passed a string that is known
2504 to be a valid IP address. However, in the case of IPv6 addresses, it was
2505 not checking this. This is a hostage to fortune. Exim now panics and dies
2506 if the condition is not met. A case was found where this could be provoked
2507 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2508 components; fortuitously, this particular loophole had already been fixed
2509 by change 4.50/55 or 4.44/33 above.
2511 If there are any other similar loopholes, the new check in host_aton()
2512 itself should stop them being exploited. The report I received stated that
2513 data on the command line could provoke the exploit when Exim was running as
2514 exim, but did not say which command line option was involved. All I could
2515 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2516 running as the user.
2519 35. There was a buffer overflow vulnerability in the SPA authentication code
2520 (which came originally from the Samba project). I have added a test to the
2521 spa_base64_to_bits() function which I hope fixes it.
2524 36. The daemon start-up calls getloadavg() while still root for those OS that
2525 need the first call to be done as root, but it missed one case: when
2526 deliver_queue_load_max is set with deliver_drop_privilege. This is
2527 necessary for the benefit of the queue runner, because there is no re-exec
2528 when deliver_drop_privilege is set.
2530 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2531 was in use, was not putting the data itself into the right store pool;
2532 consequently, it could be overwritten for a subsequent message in the same
2533 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2536 38. Sometimes the final signoff response after QUIT could fail to get
2537 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2538 < 0 before doing a fflush(). This bug looks as though it goes back to the
2539 introduction of TLS in release 3.20, but "sometimes" must have been rare
2540 because the tests only now provoked it.
2546 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2547 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2548 Now it returns an integer. A number of calls were still expecting a T/F
2549 return. Fortuitously, in all cases, the tests worked in OK situations,
2550 which is the norm. However, things would have gone wrong on any write
2551 failures on the smtp file descriptor. This function is used when sending
2552 messages over SMTP and also when doing verify callouts.
2554 2. When Exim is called to do synchronous delivery of a locally submitted
2555 message (the -odf or -odi options), it no longer closes stderr before doing
2558 3. Implemented the mua_wrapper option.
2560 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2562 5. Implemented the functions header_remove(), header_testname(),
2563 header_add_at_position(), and receive_remove_recipient(), and exported them
2566 6. If an ACL "warn" statement specified the addition of headers, Exim already
2567 inserted X-ACL-Warn: at the start if there was no header name. However, it
2568 was not making this test for the second and subsequent header lines if
2569 there were newlines in the string. This meant that an invalid header could
2570 be inserted if Exim was badly configured.
2572 7. Allow an ACL "warn" statement to add header lines at the start or after all
2573 the Received: headers, as well as at the end.
2575 8. Added the rcpt_4xx retry error code.
2577 9. Added postmaster_mailfrom=xxx to callout verification option.
2579 10. Added mailfrom=xxxx to the callout verification option, for verify=
2582 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2583 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2585 12. Inserted some casts to stop certain compilers warning when using pointer
2586 differences as field lengths or precisions in printf-type calls (mostly
2587 affecting debugging statements).
2589 13. Added optional readline() support for -be (dynamically loaded).
2591 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2592 same clock tick as a message's arrival, so that its received time was the
2593 same as the "first fail" time on the retry record, and that message
2594 remained on the queue past the ultimate address timeout, every queue runner
2595 would try a delivery (because it was past the ultimate address timeout) but
2596 after another failure, the ultimate address timeout, which should have then
2597 bounced the address, did not kick in. This was a "< instead of <=" error;
2598 in most cases the first failure would have been in the next clock tick
2599 after the received time, and all would be well.
2601 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2602 being recognized when the domain list was tested by the match_domain
2603 condition in an expansion string.
2605 16. Added the ${str2b64: operator.
2607 17. Exim was always calling setrlimit() to set a large limit for the number of
2608 processes, without checking whether the existing limit was already
2609 adequate. (It did check for the limit on file descriptors.) Furthermore,
2610 errors from getrlimit() and setrlimit() were being ignored. Now they are
2611 logged to the main and panic logs, but Exim does carry on, to try to do its
2612 job under whatever limits there are.
2614 18. Imported PCRE 5.0.
2616 19. Trivial typo in log message " temporarily refused connection" (the leading
2619 20. If the log selector return_path_on_delivery was set and an address was
2620 redirected to /dev/null, the delivery process crashed because it assumed
2621 that a return path would always be set for a "successful" delivery. In this
2622 case, the whole delivery is bypassed as an optimization, and therefore no
2625 21. Internal re-arrangement: the function for sending a challenge and reading
2626 a response while authentication was assuming a zero-terminated challenge
2627 string. It's now changed to take a pointer and a length, to allow for
2628 binary data in such strings.
2630 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2632 23. Exim was not respecting finduser_retries when seeking the login of the
2633 uid under which it was called; it was always trying 10 times. (The default
2634 setting of finduser_retries is zero.) Also, it was sleeping after the final
2635 failure, which is pointless.
2637 24. Implemented tls_on_connect_ports.
2639 25. Implemented acl_smtp_predata.
2641 26. If the domain in control=submission is set empty, Exim assumes that the
2642 authenticated id is a complete email address when it generates From: or
2643 Sender: header lines.
2645 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2646 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2647 chown and chgrp in /bin and hostname in /usr/bin.
2649 28. Exim was keeping the "process log" file open after each use, just as it
2650 does for the main log. This opens the possibility of it remaining open for
2651 long periods when the USR1 signal hits a daemon. Occasional processlog
2652 errors were reported, that could have been caused by this. Anyway, it seems
2653 much more sensible not to leave this file open at all, so that is what now
2656 29. The long-running daemon process does not normally write to the log once it
2657 has entered its main loop, and it closes the log before doing so. This is
2658 so that log files can straightforwardly be renamed and moved. However,
2659 there are a couple of unusual error situations where the daemon does write
2660 log entries, and I had neglected to close the log afterwards.
2662 30. The text of an SMTP error response that was received during a remote
2663 delivery was being truncated at 512 bytes. This is too short for some of
2664 the long messages that one sometimes sees. I've increased the limit to
2667 31. It is now possible to make retry rules that apply only when a message has a
2668 specific sender, in particular, an empty sender.
2670 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2671 it possible to be selective about when SMTP synchronization is enforced.
2673 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2675 32. Implemented hosts_connection_nolog.
2677 33. Added an ACL for QUIT.
2679 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2682 35. Added mailbox_size and mailbox_filecount to appendfile.
2684 36. Added control = no_multiline_responses to ACLs.
2686 37. There was a bug in the logic of the code that waits for the clock to tick
2687 in the case where the clock went backwards by a substantial amount such
2688 that the microsecond fraction of "now" was more than the microsecond
2689 fraction of "then" (but the whole seconds number was less).
2691 38. Added support for the libradius Radius client library this is found on
2692 FreeBSD (previously only the radiusclient library was supported).
2698 1. When certain lookups returned multiple values in the form name=value, the
2699 quoting of the values was not always being done properly. Specifically:
2700 (a) If the value started with a double quote, but contained no whitespace,
2702 (b) If the value contained whitespace other than a space character (i.e.
2703 tabs or newlines or carriage returns) it was not quoted.
2704 This fix has been applied to the mysql and pgsql lookups by writing a
2705 separate quoting function and calling it from the lookup code. The fix
2706 should probably also be applied to nisplus, ibase and oracle lookups, but
2707 since I cannot test any of those, I have not disturbed their existing code.
2709 2. A hit in the callout cache for a specific address caused a log line with no
2710 reason for rejecting RCPT. Now it says "Previous (cached) callout
2711 verification failure".
2713 3. There was an off-by-one bug in the queryprogram router. An over-long
2714 return line was truncated at 256 instead of 255 characters, thereby
2715 overflowing its buffer with the terminating zero. As well as fixing this, I
2716 have increased the buffer size to 1024 (and made a note to document this).
2718 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2719 when Exim is waiting for an SMTP response from a remote server, Exim
2720 restarts its select() call on the socket, thereby resetting its timeout.
2721 This is not a problem when such interrupts are rare. Somebody set up a cron
2722 job to run exiwhat every 2 minutes, which is less than the normal select()
2723 timeout (5 or 10 minutes). This meant that the select() timeout never
2724 kicked in because it was always reset. I have fixed this by comparing the
2725 time when an interrupt arrives with the time at the start of the first call
2726 to select(). If more time than the timeout has elapsed, the interrupt is
2727 treated as a timeout.
2729 5. Some internal re-factoring in preparation for the addition of Sieve
2730 extensions (by MH). In particular, the "personal" test is moved to a
2731 separate function, and given an option for scanning Cc: and Bcc: (which is
2732 not set for Exim filters).
2734 6. When Exim created an email address using the login of the caller as the
2735 local part (e.g. when creating a From: or Sender: header line), it was not
2736 quoting the local part when it contained special characters such as @.
2738 7. Installed new OpenBSD configuration files.
2740 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2741 try to make them clearer.
2743 9. Callout options, other than the timeout value, were being ignored when
2744 verifying sender addresses in header lines. For example, when using
2746 verify = header_sender/callout=no_cache
2748 the cache was (incorrectly) being used.
2750 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2751 only the Cygwin environment.
2753 11. When return_path_on_delivery was set as a log selector, if different remote
2754 addresses in the same message used different return paths and parallel
2755 remote delivery occurred, the wrong values would sometimes be logged.
2756 (Whenever a remote delivery process finished, the return path value from
2757 the most recently started remote delivery process was logged.)
2759 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2760 lines when AUTH and/or TLS are in use. This is the "received protocol"
2761 field. Exim used to use "asmtp" for authenticated SMTP, without any
2762 indication (in the protocol name) for TLS use. Now it follows the RFC and
2763 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2764 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2765 names appear in log lines as well as in Received: header lines.
2767 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2768 extensions, and comparison tests, and to fix some bugs.
2770 14. Changes to the "personal" filter test:
2772 (1) The test was buggy in that it was just doing the equivalent of
2773 "contains" tests on header lines. For example, if a user's address was
2774 anne@some.where, the "personal" test would incorrectly be true for
2776 To: susanne@some.where
2778 This test is now done by extracting each address from the header in turn,
2779 and checking the entire address. Other tests that are part of "personal"
2780 are now done using regular expressions (for example, to check local parts
2781 of addresses in From: header lines).
2783 (2) The list of non-personal local parts in From: addresses has been
2784 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2785 taken from the Sieve specification recommendations.
2787 (3) If the message contains any header line starting with "List-" it is
2788 treated as non-personal.
2790 (4) The test for "circular" in the Subject: header line has been removed
2791 because it now seems ill-conceived.
2793 15. Minor typos in src/EDITME comments corrected.
2795 16. Installed latest exipick from John Jetmore.
2797 17. If headers_add on a router specified a text string that was too long for
2798 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2799 of string_sprintf() is now avoided.
2801 18. $message_body_size was not set (it was always zero) when running the DATA
2802 ACL and the local_scan() function.
2804 19. For the "mail" command in an Exim filter, no default was being set for
2805 the once_repeat time, causing a random time value to be used if "once" was
2806 specified. (If the value happened to be <= 0, no repeat happened.) The
2807 default is now 0s, meaning "never repeat". The "vacation" command was OK
2808 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2809 (I found it when inspecting the code).
2811 20. There is now an overall timeout for performing a callout verification. It
2812 defaults to 4 times the callout timeout, which applies to individual SMTP
2813 commands during the callout. The overall timeout applies when there is more
2814 than one host that can be tried. The timeout is checked before trying the
2815 next host. This prevents very long delays if there are a large number of
2816 hosts and all are timing out (e.g. when the network connections are timing
2817 out). The value of the overall timeout can be changed by specifying an
2818 additional sub-option for "callout", called "maxwait". For example:
2820 verify = sender/callout=5s,maxwait=20s
2822 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2823 to the end before writing, but this should make it even safer).
2825 22. Exim was forgetting that it had advertised PIPELINING for the second and
2826 subsequent messages on an SMTP connection. It was also not resetting its
2827 memory on STARTTLS and an internal HELO.
2829 23. When Exim logs an SMTP synchronization error within a session, it now
2830 records whether PIPELINING has been advertised or not.
2832 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2833 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2834 rather than long int.
2836 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2838 26. Added the never_mail option to autoreply.
2844 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2845 crash if the getsockname() call failed; this can happen if a connection is
2846 closed very soon after it is established. The problem was simply in the
2847 order in which certain operations were done, causing Exim to try to write
2848 to the SMTP stream before it had set up the file descriptor. The bug has
2849 been fixed by making things happen in the correct order.
2855 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2856 before the connection was closed, thus losing the rejection response.
2858 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2859 some early Solaris releases, but causes trouble in current releases where
2860 socklen_t is defined.
2862 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2865 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2868 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2869 that releasing the top of it at the end releases what was used for sub-
2870 expansions (unless the block got too big). However, discard this block if
2871 the first thing is a variable or header, so that we can use its block when
2872 it is dynamic (useful for very large $message_headers, for example).
2874 6. Lookups now cache *every* query, not just the most recent. A new, separate
2875 store pool is used for this. It can be recovered when all lookup caches are
2876 flushed. Lookups now release memory at the end of their result strings.
2877 This has involved some general refactoring of the lookup sources.
2879 7. Some code has been added to the store_xxx() functions to reduce the amount
2880 of flapping under certain conditions.
2882 8. log_incoming_interface used to affect only the <= reception log lines. Now
2883 it causes the local interface and port to be added to several more SMTP log
2884 lines, for example "SMTP connection from", and rejection lines.
2886 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2888 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2890 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2891 could be overwritten at the end of the current message (or the start of a
2892 new message if it was set in a HELO ACL). The value is now preserved for
2893 the duration of the SMTP connection.
2895 12. If a transport had a headers_rewrite setting, and a matching header line
2896 contained an unqualified address, that address was qualified, even if it
2897 did not match any rewriting rules. The underlying bug was that the values
2898 of the flags that permit the existence of unqualified sender and recipient
2899 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2900 non-local messages, and by -bnq for local messages) were not being
2901 preserved with the message after it was received.
2903 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2904 "next input=" as part of the text comprising the host identity instead of
2905 the correct text. The code was using the same buffer for two different
2906 strings. However, depending on which order the printing function evaluated
2907 its arguments, the bug did not always show up. Under Linux, for example, my
2908 test suite worked just fine.
2910 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2911 doesn't work with some older versions of Perl. It has been changed to "my",
2912 which in any case is probably the better facility to use.
2914 15. A really picky compiler found some instances of statements for creating
2915 error messages that either had too many or two few arguments for the format
2918 16. The size of the buffer for calls to the DNS resolver has been increased
2919 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2920 for addresses that have a lot of PTR records. This alleviates a problem; it
2921 does not fully solve it.
2923 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2924 buffer now truncates the list and logs the incident, which is the same
2925 action as happens when Exim is looking up a host name and its aliases.
2926 Previously in this situation something unpredictable would happen;
2927 sometimes it was "internal error: store_reset failed".
2929 18. If a server dropped the connection unexpectedly when an Exim client was
2930 using GnuTLS and trying to read a response, the client delivery process
2931 crashed while trying to generate an error log message.
2933 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2934 string, for example:
2936 warn message = H1: something\nH2: something
2938 the text was added as a single header line from Exim's point of view
2939 though it ended up OK in the delivered message. However, searching for the
2940 second and subsequent header lines using $h_h2: did not work. This has been
2941 fixed. Similarly, if a system filter added multiple headers in this way,
2942 the routers could not see them.
2944 20. Expanded the error message when iplsearch is called with an invalid key to
2945 suggest using net-iplsearch in a host list.
2947 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2948 ACLs are no longer actually imposed (and a message to that effect is
2951 22. If a "gecos" field in a passwd entry contained escaped characters, in
2952 particular, if it contained a \" sequence, Exim got it wrong when building
2953 a From: or a Sender: header from that name. A second bug also caused
2954 incorrect handling when an unquoted " was present following a character
2955 that needed quoting.
2957 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2958 was not being matched caselessly.
2960 24. Arranged for all hyphens in the exim.8 source to be escaped with
2963 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2964 properly. Recipient callout cache records were still being keyed to include
2965 the sender, even when use_sender was set false. This led to far more
2966 callouts that were necessary. The sender is no longer included in the key
2967 when use_sender is false.
2969 26. Added "control = submission" modifier to ACLs.
2971 27. Added the ${base62d: operator to decode base 62 numbers.
2973 28. dnsdb lookups can now access SRV records.
2975 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2976 the configuration file.
2978 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2979 (-v) mode. This makes the output for a verbose queue run more intelligible.
2981 31. Added a use_postmaster feature to recipient callouts.
2983 32. Added the $body_zerocount variable, containing the number of binary zero
2984 bytes in the message body.
2986 33. The time of last modification of the "new" subdirectory is now used as the
2987 "mailbox time last read" when there is a quota error for a maildir
2990 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2992 35. Added +ignore_unknown as a special item in host lists.
2994 36. Code for decoding IPv6 addresses in host lists is now included, even if
2995 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2996 address was recognized as an IP address, but was then not correctly decoded
2997 into binary, causing unexpected and incorrect effects when compared with
3004 1. Very minor rewording of debugging text in manualroute to say "list of
3005 hosts" instead of "hostlist".
3007 2. If verify=header_syntax was set, and a header line with an unqualified
3008 address (no domain) and a large number of spaces between the end of the
3009 name and the colon was received, the reception process suffered a buffer
3010 overflow, and (when I tested it) crashed. This was caused by some obsolete
3011 code that should have been removed. The fix is to remove it!
3013 3. When running in the test harness, delay a bit after writing a bounce
3014 message to get a bit more predictability in the log output.
3016 4. Added a call to search_tidyup() just before forking a reception process. In
3017 theory, someone could use a lookup in the expansion of smtp_accept_max_
3018 per_host which, without the tidyup, could leave open a database connection.
3020 5. Added the variables $recipient_data and $sender_data which get set from a
3021 lookup success in an ACL "recipients" or "senders" condition, or a router
3022 "senders" option, similar to $domain_data and $local_part_data.
3024 6. Moved the writing of debug_print from before to after the "senders" test
3027 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3028 problems for message scanning, either using a data ACL, or using
3029 local_scan() because the Received: header was not generated till after they
3030 were called (in order to set the time as the time of reception completion).
3031 I have revised the way this works. The header is now generated after the
3032 body is received, but before the ACL or local_scan() are called. After they
3033 are run, the timestamp in the header is updated.
3039 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3040 before starting a queue runner without re-exec. This happened only when
3041 deliver_drop_privilege was set or when the Exim user was set to root. The
3042 effect of the bug was that timeouts during subsequent deliveries caused
3043 crashes instead of being properly handled. The handler is now left at its
3044 default (and expected) setting.
3046 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3047 message, again when deliver_drop_privilege is set or Exim is run as root.
3048 The bug described in (1) was not present in this case, but the tidying up
3049 of the other signals was missing. I have made the two cases consistent.
3051 3. The ignore_target_hosts setting on a manualroute router was being ignored
3052 for hosts that were looked up using the /MX notation.
3054 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3057 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3058 operated on the sender address. After changing the $sender_address to <>
3059 for the sender address verify, Exim was re-instated it as the original
3060 (before rewriting) address, but remembering that it had rewritten it, so it
3061 wasn't rewriting it again. This bug also had the effect of breaking the
3062 sender address verification caching when the sender address was rewritten.
3064 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3065 This has been changed so that if the ip literal address matches
3066 ignore_target_hosts, the router declines.
3068 7. Added expansion conditions match_domain, match_address, and match_local_
3069 part (NOT match_host).
3071 8. The placeholder for the Received: header didn't have a length field set.
3073 9. Added code to Exim itself and to exim_lock to test for a specific race
3074 condition that could lead to file corruption when using MBX delivery. The
3075 issue is with the lockfile that is created in /tmp. If this file is removed
3076 after a process has opened it but before that process has acquired a lock,
3077 there is the potential for a second process to recreate the file and also
3078 acquire a lock. This could lead to two Exim processes writing to the file
3079 at the same time. The added code performs the same test as UW imapd; it
3080 checks after acquiring the lock that its file descriptor still refers to
3081 the same named file.
3083 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3084 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3087 11. Added the smtp_active_hostname option. If used, this will typically be made
3088 to depend on the incoming interface address. Because $interface_address is
3089 not set up until the daemon has forked a reception process, error responses
3090 that can happen earlier (such as "too many connections") no longer contain
3093 12. If an expansion in a condition on a "warn" statement fails because a lookup
3094 defers, the "warn" statement is abandoned, and the next ACL statement is
3095 processed. Previously this caused the whole ACL to be aborted.
3097 13. Added the iplsearch lookup type.
3099 14. Added ident_timeout as a log selector.
3101 15. Added tls_certificate_verified as a log selector.
3103 16. Added a global option tls_require_ciphers (compare the smtp transport
3104 option of the same name). This controls incoming TLS connections.
3106 17. I finally figured out how to make tls_require_ciphers do a similar thing
3107 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3108 before starting the TLS session.
3110 18. Tabs are now shown as \t in -bP output.
3112 19. If the log selector return_path_on_delivery was set, Exim crashed when
3113 bouncing a message because it had too many Received: header lines.
3115 20. If two routers both had headers_remove settings, and the first one included
3116 a superfluous trailing colon, the final name in the first list and the
3117 first name in the second list were incorrectly joined into one item (with a
3118 colon in the middle).
3124 1. Added -C and -D options to the exinext utility, mainly to make it easier
3125 to include in the automated testing, but these could be helpful when
3126 multiple configurations are in use.
3128 2. The exinext utility was not formatting the output nicely when there was
3129 an alternate port involved in the retry record key, nor when there was a
3130 message id as well (for retries that were specific to a specific message
3131 and a specific host). It was also confused by IPv6 addresses, because of
3132 the additional colons they contain. I have fixed the IPv4 problem, and
3133 patched it up to do a reasonable job for IPv6.
3135 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3136 delivery, the log line now contains "pipelined" if PIPELINING was used.
3138 4. An SMTP transport process used to panic and die if the bind() call to set
3139 an explicit outgoing interface failed. This has been changed; it is now
3140 treated in the same way as a connect() failure.
3142 5. A reference to $sender_host_name in the part of a conditional expansion
3143 that was being skipped was still causing a DNS lookup. This no longer
3146 6. The def: expansion condition was not recognizing references to header lines
3147 that used bh_ and bheader_.
3149 7. Added the _cache feature to named lists.
3151 8. The code for checking quota_filecount in the appendfile transport was
3152 allowing one more file than it should have been.
3154 9. For compatibility with Sendmail, the command line option
3162 and sets the incoming protocol and host name (for trusted callers). The
3163 host name and its colon can be omitted when only the protocol is to be set.
3164 Note the Exim already has two private options, -pd and -ps, that refer to
3165 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3166 "s", but I don't think that's a major issue.
3168 10. A number of refactoring changes to the code, none of which should affect
3171 (a) The number of logging options was getting close to filling up the
3172 32-bit word that was used as a bit map. I have split them into two classes:
3173 those that are passed in the argument to log_write(), and those that are
3174 only ever tested independently outside of that function. These are now in
3175 separate 32-bit words, so there is plenty of room for expansion again.
3176 There is no change in the user interface or the logging behaviour.
3178 (b) When building, for example, log lines, the code previously used a
3179 macro that called string_cat() twice, in order to add two strings. This is
3180 not really sufficiently general. Furthermore, there was one instance where
3181 it was actually wrong because one of the argument was used twice, and in
3182 one call a function was used. (As it happened, calling the function twice
3183 did not affect the overall behaviour.) The macro has been replaced by a
3184 function that can join an arbitrary number of extra strings onto a growing
3187 (c) The code for expansion conditions now uses a table and a binary chop
3188 instead of a serial search (which was left over from when there were very
3189 few conditions). Also, it now recognizes conditions like "pam" even when
3190 the relevant support is not compiled in: a suitably worded error message is
3191 given if an attempt is made to use such a condition.
3193 11. Added ${time_interval:xxxxx}.
3195 12. A bug was causing one of the ddress fields not to be passed back correctly
3196 from remote delivery subprocesses. The field in question was not being
3197 subsequently used, so this caused to problems in practice.
3199 13. Added new log selectors queue_time and deliver_time.
3201 14. Might have fixed a bug in maildirsizefile handling that threw up
3202 "unexpected character" debug warnings, and recalculated the data
3203 unnecessarily. In any case, I expanded the warning message to give more
3206 15. Added the message "Restricted characters in address" to the statements in
3207 the default ACL that block characters like @ and % in local parts.
3209 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3210 Three changes have been made:
3212 (a) There was a serious bug; a negative response to MAIL caused the whole
3213 recipient domain to be cached as invalid, thereby blocking all messages
3214 to all local parts at the same domain, from all senders. This bug has
3215 been fixed. The domain is no longer cached after a negative response to
3216 MAIL if the sender used is not empty.
3218 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3221 (c) A new callout option, "use_sender" has been added for people who want
3222 the modified behaviour.
3228 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3231 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3232 indeed breaks things for older releases.
3234 3. Added additional logging to the case where there is a problem reading data
3235 from a filter that is running in a subprocess using a pipe, in order to
3236 try to track down a specific problem.
3238 4. Testing facility fudge: when running in the test harness and attempting
3239 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3240 getting "No route to host". Convert this to a timeout.
3242 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3245 6. Some OS don't have socklen_t but use size_t instead. This affects the
3246 fifth argument of getsockopt() amongst other things. This is now
3247 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3248 can be set for individual OS. I have set it for SunOS5, OSF1, and
3249 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3250 some earlier ones do not.
3252 7. Change 4.30/15 was not doing the test caselessly.
3254 8. The standard form for an IPv6 address literal was being rejected by address
3255 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3256 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3257 this, as well as the form without the "IPv6" on the front (but only when
3258 address literals are enabled, of course).
3260 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3262 10. Exim crashed if a message with an empty sender address specified by -f
3263 encountered a router with an errors_to setting. This could be provoked only
3264 by a command such as
3268 where an empty string was supplied; "<>" did not hit this bug.
3270 11. Installed PCRE release 4.5.
3272 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3273 remained set. It is now erased.
3275 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3276 times from message ids (which are base 36 rather than the normal 62).
3278 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3279 were being counted as actual protocol errors, and logged if the log
3280 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3281 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3282 and DATA following a set of rejected RCPTs do not count as protocol errors.
3283 In other words, Exim assumes they were pipelined, though this may not
3284 actually be the case. Of course, in all cases the client gets an
3285 appropriate error code.
3287 15. If a lookup fails in an ACL condition, a message about the failure may
3288 be available; it is used if testing the ACL cannot continue, because most
3289 such messages specify what the cause of the deferral is. However, some
3290 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3291 that caused an old message to be retained and used if a later statement
3292 caused a defer, replacing the real cause of the deferral.
3294 16. If an IP address had so many PTR records that the DNS lookup buffer
3295 was not large enough to hold them, Exim could crash while trying to process
3296 the truncated data. It now detects and logs this case.
3298 17. Further to 4.21/58, another change has been made: if (and only if) the
3299 first line of a message (the first header line) ends with CRLF, a bare LF
3300 in a subsequent header line has a space inserted after it, so as not to
3301 terminate the header.
3303 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3304 unnecessarily, used atoi() instead of strtol(), and didn't check the
3305 termination when getting file sizes from file names by regex.
3307 19. Completely re-implemented the support for maildirsize files, in the light
3308 of a number of problems with the previous contributed implementation
3309 (4.30/29). In particular:
3311 . If the quota is zero, the maildirsize file is maintained, but no quota is
3314 . If the maildir directory does not exist, it is created before any attempt
3315 to write a maildirsize file.
3317 . The quota value in the file is just a cache; if the quota is changed in
3318 the transport, the new value overrides.
3320 . A regular expression is available for excluding directories from the
3323 20. The autoreply transport checks the characters in options that define the
3324 message's headers; it allows continued headers, but it was checking with
3325 isspace() after an embedded newline instead of explicitly looking for a
3328 21. If all the "regular" hosts to which an address was routed had passed their
3329 expiry times, and had not reached their retry times, the address was
3330 bounced, even if fallback hosts were defined. Now Exim should go on to try
3333 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3334 equivalent code in the SMTP transport. Some hosts send humungous responses
3335 to HELO/EHLO, more than 1024 it seems.
3337 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3338 gives compiler warnings in some environments. I've now done it "properly",
3341 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3342 (because of problems with the built-in one) was declared to return uschar *
3343 instead of char *, causing compiler failure.
3345 25. Fixed a file descriptor leak when processing alias/forward files.
3347 26. Fixed a minor format string issue in dbfn.c.
3349 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3351 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3352 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3355 29. When a sender address is verified, it is cached, to save repeating the test
3356 when there is more than one recipient in a message. However, when the
3357 verification involves a callout, it is possible for different callout
3358 options to be set for different recipients. It is too complicated to keep
3359 track of this in the cache, so now Exim always runs a verification when a
3360 callout is required, relying on the callout cache for the optimization.
3361 The overhead is duplication of the address routing, but this should not be
3364 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3365 to be verified with callout=postmaster, and the main callout worked but the
3366 postmaster check failed, the verification correctly failed. However, if a
3367 subsequent RCPT command asked for sender verification *without* the
3368 postmaster check, incorrect caching caused this verification also to fail,
3371 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3372 it was not caching the DNS options (qualify_single, search_parents) that
3373 were used when the lookup failed. A subsequent lookup with different
3374 options therefore always gave the same answer, though there were cases
3375 where it should not have. (Example: a "domains = !$mx_any" option on a
3376 dnslookup router: the "domains" option is always processed without any
3377 widening, but the router might have qualify_single set.) Now Exim uses the
3378 cached value only when the same options are set.
3380 32. Added John Jetmore's "exipick" utility to the distribution.
3382 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3383 than a timeout (e.g. a certificate is required, and is not provided), an
3384 Exim server now closes the connection immediately. Previously it waited for
3385 the client to close - but if the client is SSL, it seems that they each
3386 wait for each other, leading to a delay before one of them times out.
3388 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3389 maintained 0.8.x compatibility because I don't think many are using it, and
3390 it is clearly obsolete.
3392 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3395 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3396 client certificate was expired. A simple patch fixes this, though I don't
3397 understand the full logic of why the verify callback is called multiple
3400 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3401 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3402 which causes problems with some clients (such as the Certicom SSL Plus
3403 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3404 disables the coutermeasure allowing Eudora to connect."
3406 38. Exim was not checking that a write() to a log file succeeded. This could
3407 lead to Bad Things if a log got too big, in particular if it hit a file
3408 size limit. Exim now panics and dies if it cannot write to a log file, just
3409 as it does if it cannot open a log file.
3411 39. Modified OS/Makefile-Linux so that it now contains
3413 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3415 The two -D definitions ensure that Exim is compiled with large file
3416 support, which makes it possible to handle log files that are bigger than
3419 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3420 instance) a domain was checked against a named list that involved a lookup,
3421 causing $domain_data to be set, then another domain was checked against the
3422 same list, then the first domain was re-checked, the value of $domain_data
3423 after the final check could be wrong. In particular, if the second check
3424 failed, it could be set empty. This bug probably also applied to
3427 41. The strip_trailing_dot option was not being applied to the address given
3428 with the -f command-line option.
3430 42. The code for reading a message's header from the spool was incrementing
3431 $received_count, but never initializing it. This meant that the value was
3432 incorrect (doubled) while delivering a message in the same process in which
3433 it was received. In the most common configuration of Exim, this never
3434 happens - a fresh exec is done - but it can happen when
3435 deliver_drop_privilege is set.
3437 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3438 it now includes up to 150 characters of the unexpected data in the log
3441 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3442 and building data strings. The size of both of these buffers was 10 000
3443 bytes - far larger than anybody would *ever* want, thought I. Needless to
3444 say, somebody hit the limit. I have increased the maximum line length to
3445 20 000 and the maximum data length of concatenated lines to 100 000. I have
3446 also fixed two bugs, because there was no checking on these buffers. Tsk,
3447 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3448 buffer is too small.
3450 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3451 lsearch lookups. Now it does.
3453 46. When parsing a route_list item in a manualroute router, a fixed-length
3454 buffer was used for the list of hosts. I made this 1024 bytes long,
3455 thinking that nobody would ever have a list of hosts that long. Wrong.
3456 Somebody had a whole pile of complicated expansion conditions, and the
3457 string was silently truncated, leading to an expansion error. It turns out
3458 that it is easier to change to an unlimited length (owing to other changes
3459 that have happened since this code was originally written) than to build
3460 structure for giving a limitation error. The length of the item that
3461 expands into the list of hosts is now unlimited.
3463 47. The lsearch lookup could not handle data where the length of text line was
3464 more than 4095 characters. Such lines were truncated, leading to shortened
3465 data being returned. It should now handle lines of any length.
3467 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3468 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3471 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3472 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3473 no longer happens in comment lines. A list of replacements is now placed
3474 at the head of all of the source files, except those whose only change is
3475 to replace PERL_COMMAND in the very first #! line.
3477 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3478 messages on the queue, with a bottom-up merge sort, using code contributed
3479 by Michael Haardt. This should make operations like -bp somewhat faster on
3480 large queues. It won't affect queue runners, except when queue_run_in_order
3483 51. Installed eximstats 1.31 in the distribution.
3485 52. Added support for SRV lookups to the dnslookup router.
3487 53. If an ACL referred to $message_body or $message_body_end, the value was not
3488 reset for any messages that followed in the same SMTP session.
3490 54. The store-handling optimization for building very long strings was not
3491 differentiating between the different store pools. I don't think this
3492 actually made any difference in practice, but I've tidied it.
3494 55. While running the routers to verify a sender address, $sender_address
3495 was still set to the sender address. This is wrong, because when routing to
3496 send a bounce to the sender, it would be empty. Therefore, I have changed
3497 it so that, while verifying a sender address, $sender_address is set to <>.
3498 (There is no change to what happens when verifying a recipient address.)
3500 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3501 target A or AAAA records (if not already returned) without resetting the
3502 qualify_single or search_parents options of the DNS resolver. These are
3503 inappropriate in this case because the targets of MX and SRV records must
3504 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3505 target that, when qualified, matched something in the local domain. These
3506 two options are now turned off when doing these lookups.
3508 57. It seems that at least some releases of Reiserfs (which does not have the
3509 concept of a fixed number of inodes) returns zero and not -1 for the
3510 number of available inodes. This interacted badly with check_spool_inodes,
3511 which assumed that -1 was the "no such thing" setting. What I have done is
3512 to check that the total number of inodes is greater than zero before doing
3513 the test of how many are available.
3515 58. When a "warn" ACL statement has a log_message modifier, the message is
3516 remembered, and not repeated. This is to avoid a lot of repetition when a
3517 message has many recipients that cause the same warning to be written.
3518 Howewer, Exim was preserving the list of already written lines for an
3519 entire SMTP session, which doesn't seem right. The memory is now reset if a
3520 new message is started.
3522 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3523 debugging output unless log_rewrite was also set.
3525 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3526 of (int)(handle) when we know that handle contains (void *)(-1).
3528 61. The Exim daemon panic-logs an error return when it closes the incoming
3529 connection. However "connection reset by peer" seems to be common, and
3530 isn't really an error worthy of noting specially, so that particular error
3533 62. When Exim is trying to find all the local interfaces, it used to panic and
3534 die if the ioctl to get the interface flags failed. However, it seems that
3535 on at least one OS (Solaris 9) it is possible to have an interface that is
3536 included in the list of interfaces, but for which you get a failure error
3537 for this call. This happens when the interface is not "plumbed" into a
3538 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3539 failure of the "get flags" call assumes that the interface is down.
3541 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3542 makes life easier for people who are doing arithmetic on fields extracted
3543 from dates, where you often get leading zeros that should not be
3544 interpreted as octal.
3546 64. Added qualify_domain to the redirect router, to override the global
3549 65. If a pathologically long header line contained very many addresses (the
3550 report of this problem mentioned 10 000) and each of them was rewritten,
3551 Exim could use up a very large amount of memory. (It kept on making new
3552 copies of the header line as it rewrote, and never released the old ones.)
3553 At the expense of a bit more processing, the header rewriting function has
3554 been changed so that it no longer eats memory in this way.
3556 66. The generation of the Received: header has been moved from the time that a
3557 message starts to be received, to the time that it finishes. The timestamp
3558 in the Received: header should now be very close to that of the <= log
3559 line. There are two side-effects of this change:
3561 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3562 logged header lines no longer include the local Received: line, because
3563 it has not yet been created. The same applies to a copy of the message
3564 that is returned to a non-SMTP sender when a message is rejected.
3566 (b) When a filter file is tested using -bf, no additional Received: header
3567 is added to the test message. After some thought, I decided that this
3570 This change does not affect the value of $received_for. It is still set
3571 after address rewriting, but before local_scan() is called.
3573 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3575 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3576 gave an unhelpful panic error message, and a defer error. I have managed to
3577 change this behaviour so that it now rejects any supplied certificate,
3578 which seems right, as the list of acceptable certificates is empty.
3580 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3581 gave an unhelpful defer error. I have not managed to make this reject any
3582 supplied certificates, but the error message it gives is "no certificate
3583 supplied", which is not helpful.
3585 70. exigrep's output now also includes lines that are not associated with any
3586 message, but which match the given pattern. Implemented by a patch from
3587 Martin Sluka, which also tidied up the Perl a bit.
3589 71. Recipient callout verification, like sender verification, was using <> in
3590 the MAIL FROM command. This isn't really the right thing, since the actual
3591 sender may affect whether the remote host accepts the recipient or not. I
3592 have changed it to use the actual sender in the callout; this means that
3593 the cache record is now keyed on a recipient/sender pair, not just the
3594 recipient address. There doesn't seem to be a real danger of callout loops,
3595 since a callout by the remote host to check the sender would use <>.
3596 [SEE ABOVE: changed after hitting problems.]
3598 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3599 temporary errors. However, in the case of such a code being given after
3600 the end of a data transmission (i.e. after ".") Exim was failing to write
3601 a retry record for the message. (Yes, there was some broken host that was
3602 actually sending 8xx at this point.)
3604 73. An unknown lookup type in a host list could cause Exim to panic-die when
3605 the list was checked. (An example that provoked this was putting <; in the
3606 middle of a list instead of at the start.) If this happened during a DATA
3607 ACL check, a -D file could be left lying around. This kind of configuration
3608 error no longer causes Exim to die; instead it causes a defer errror. The
3609 incident is still logged to the main and panic logs.
3611 74. Buglet left over from Exim 3 conversion. The message "too many messages
3612 in one connection" was written to the rejectlog but not the mainlog, except
3613 when address rewriting (yes!) was being logged.
3615 75. Added write_rejectlog option.
3617 76. When a system filter was run not as root (that is, when system_filter_user
3618 was set), the values of the $n variables were not being returned to the
3619 main process; thus, they were not subsequently available in the $sn
3622 77. Added +return_path_on_delivery log selector.
3624 78. A connection timeout was being treated differently from recipients deferred
3625 when testing hosts_max_try with a message that was older than the host's
3626 retry timeout. (The host should not be counted, thus allowing all hosts to
3627 be tried at least once before bouncing.) This may have been the cause of an
3628 occasionally reported bug whereby a message would remain on the queue
3629 longer than the retry timeout, but would be bounced if a delivery was
3630 forced. I say "may" because I never totally pinned down the problem;
3631 setting up timeout/retry tests is difficult. See also the next item.
3633 79. The ultimate address timeout was not being applied to errors that involved
3634 a combination of host plus message (for example, a timeout on a MAIL
3635 command). When an address resolved to a number of possible hosts, and they
3636 were not all tried for each delivery (e.g. because of hosts_max_try), a
3637 message could remain on the queue longer than the retry timeout.
3639 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3642 81. Fixed an obscure SMTP outgoing bug which required at least the following
3643 conditions: (a) there was another message waiting for the same server;
3644 (b) the server returned 5xx to all RCPT commands in the first message so
3645 that the message was not completed; (c) the server dropped the connection
3646 or gave a negative response to the RSET that Exim sends to abort the
3647 transaction. The observed case was a dropped connection after DATA that had
3648 been sent in pipelining mode. That is, the server had advertised PIPELINING
3649 but was not implementing it correctly. The effect of the bug was incorrect
3650 behaviour, such as trying another host, and this could lead to a crash.
3656 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3657 and daemon.c were passed as pointers to ints; they should have been
3658 pointers to socklen_t variables (which are typically unsigned ints).
3660 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3663 3. Fixed a really odd bug that affected only the testing scheme; patching a
3664 certain fixed string in the binary changed the value of another string that
3665 happened to be identical to the end of the original first string.
3667 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3668 name", it returns that address as the IP address. On some operating
3669 systems (e.g. Solaris), it also passes back the IP address string as the
3670 "host name". However, on others (e.g. Linux), it passes back an empty
3671 string. Exim wasn't checking for this, and was changing the host name to an
3672 empty string, assuming it had been canonicized.
3674 5. Although rare, it is permitted to have more than one PTR record for a given
3675 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3676 all the names associated with an address, because they do in Solaris.
3677 However, it seems that they do not in Linux for data that comes from the
3678 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3679 I found this out when I moved to a new Linux workstation and tried to run
3680 the Exim test suite.
3682 To get round this problem I have changed the code so that it now does its
3683 own call to the DNS to look up PTR records when searching for a host name.
3684 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3685 addresses that are only in /etc/hosts are still found.
3687 This behaviour is, however, controlled by an option called host_lookup_
3688 order, which defaults to "bydns:byaddr". If people want to use the other
3689 order, or indeed, just use one or the other means of lookup, they can
3690 specify it in this variable.
3692 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3693 some operating systems, this comes back from gethostbyaddr() as an empty
3694 string, and this is what Exim used to test for. However, it seems that in
3695 other systems, "." is yielded. Exim now tests for this case too.
3697 7. The values of check_spool_space and check_log_space are now held internally
3698 as a number of kilobytes instead of an absolute number of bytes. If a
3699 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3700 kilobyte. This means that much larger values can be stored.
3702 8. Exim monitor: an attempt to get the action menu when not actually pointing
3703 at a message produces an empty menu entitled "No message selected". This
3704 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3705 no entries in it ("Shell widget menu has zero width and/or height"). So I
3706 have added a single, blank menu entry in this case.
3708 9. Added ${quote_local_part.
3710 10. MIME decoding is now applied to the contents of Subject: header lines when
3713 11. Now that a reference to $sender_host_address automatically causes a reverse
3714 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3715 host lookup before query-style lookups in lists that might use this
3716 variable. This has therefore been abolished, and the "net-" prefix is no
3717 longer necessary for query-style lookups.
3719 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3720 have been a typo for LFLAGS, so it has been changed.
3722 13. The install script calls Exim with "-C /dev/null" in order to find the
3723 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3724 to be output. Howeve, since Exim outputs its version number before the
3725 error, it didn't break the script. It just looked ugly. I fixed this by
3726 always allowing "-C /dev/null" if the caller is root.
3728 14. Ignore overlarge ACL variable number when reading spool file - insurance
3729 against a later release with more variables having written the file.
3731 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3732 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3733 this, as well as the form without the "IPv6" on the front.
3735 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3736 OS/Makefile-Darwin file.
3738 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3739 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3741 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3742 variable. However, some OS use ints for inodes. I've added cast to long int
3743 to get rid of the compiler warning.
3745 19. I had forgotten to lock out "/../" in configuration file names when
3746 ALT_CONFIG_PREFIX was set.
3748 20. Routers used for verification do not need to specify transports. However,
3749 if such a router generated a host list, and callout was configured, Exim
3750 crashed, because it could not find a port number from the (non-existent)
3751 transport. It now assumes port 25 in this circumstance.
3753 21. Added the -t option to exigrep.
3755 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3756 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3757 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3758 them is set, LOOKUP_LSEARCH is forced.
3760 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3762 24. Added sender and host information to the "rejected by local_scan()" log
3763 line; previously there was no indication of these.
3765 25. Added .include_if_exists.
3767 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3768 Linux. It turns out that not all file systems support this. Apparently some
3769 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3770 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3771 not supported on the file descriptor, is now ignored when Exim is trying to
3772 sync a directory. This applies only to Linux.
3774 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3776 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3777 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3780 29. Added support for maildirsize files from supplied patch (modified a bit).
3782 30. The use of :fail: followed by an empty string could lead Exim to respond to
3783 sender verification failures with (e.g.):
3785 550 Verification failed for <xxx>
3786 550 Sender verify failed
3788 where the first response line was missing the '-' that indicates it is not
3789 the final line of the response.
3791 31. The loop for finding the name of the user that called Exim had a hardwired
3792 limit of 10; it now uses the value of finduser_retries, which is used for
3793 all other user lookups.
3795 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3798 33. Exim was neglecting to zero errno before one call of strtol() when
3799 expanding a string and expecting an integer value. On some systems this
3800 resulted in spurious "integer overflow" errors. Also, it was casting the
3801 result into an int without checking.
3803 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3804 did not work. The code looks as if it has *never* worked, though it appears
3805 to have been documented since at least releast 1.62. I have made it work.
3807 35. The "timeout_DNS" error in retry rules, also documented since at least
3808 1.62, also never worked. As it isn't clear exactly what this means, and
3809 clearly it isn't a major issue, I have abolished the feature by treating it
3810 as "timeout", and writing a warning to the main and panic logs.
3812 36. The display of retry rules for -brt wasn't always showing the error code
3815 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3816 timeout_connect_A, timeout_connect_MX.
3818 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3819 to the empty sender.
3821 39. The daemon was not analysing the content of -oX till after it had closed
3822 stderr and disconnected from the controlling terminal. This meant that any
3823 syntax errors were only noted on the panic log, and the return code from
3824 the command was 0. By re-arranging the code a little, I've made the
3825 decoding happen first, so such errors now appear on stderr, and the return
3826 code is 1. However, the actual setting up of the sockets still happens in
3827 the disconnected process, so errors there are still only recorded on the
3830 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3831 connections (as happens on some IP stacks) was logged at start up time as
3832 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3833 it from "IPv6 and IPv4", which means that two separate sockets are being
3836 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3837 says whether AF_INET or AF_INET6 was passed as an argument.
3839 42. Exiwhat output was messed up when time zones were included in log
3842 43. Exiwhat now gives more information about the daemon's listening ports,
3843 and whether -tls-on-connect was used.
3845 44. The "port" option of the smtp transport is now expanded.
3847 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3848 silently ignored. Now an error message is written to the main and panic
3851 46. There's a new ACL modifier called "logwrite" which writes to a log file
3852 as soon as it is encountered.
3854 47. Added $local_user_uid and $local_user_gid at routing time.
3856 48. Exim crashed when trying to verify a sender address that was being
3859 49. Exim was recognizing only a space character after ".include". It now also
3860 recognizes a tab character.
3862 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3863 extracting the relevant information from the specification. The man page no
3864 longer contains scrambled data for the -d option, and I've added a section
3865 at the front about calling Exim under different names.
3867 51. Added "extra_headers" argument to the "mail" command in filter files.
3869 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3872 53. Installed eximstats 1.29.
3874 54. Added transport_filter_timeout as a generic transport option.
3876 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3877 Cc: header lines. This was required by RFC 822, but it not required by RFC
3880 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3881 incoming messages that did not have them. Now it does so only if the
3882 message originates locally, that is, if there is no associated remote host
3883 address. When Resent- header lines are present, this applies to the Resent-
3884 lines rather than the non-Resent- lines.
3886 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3887 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3889 58. Messages for configuration errors now include the name of the main
3890 configuration file - useful now that there may be more than one file in a
3891 list (.included file names were always shown).
3893 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3894 for those rare installations that do not start the daemon as root or run it
3895 setuid root. I've cut out the call to initgroups() if the daemon is not
3898 60. The Exim user and group can now be bound into the binary as text strings
3899 that are looked up at the start of Exim's processing.
3901 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3903 62. Added $mailstore_basename variable.
3905 63. Installed patch to sieve.c from Michael Haardt.
3907 64. When Exim failed to open the panic log after failing to open the main log,
3908 the original message it was trying to log was written to stderr and debug
3909 output, but if they were not available (the usual case in production), it
3910 was lost. Now it is written to syslog before the two lines that record the
3911 failures to open the logs.
3913 65. Users' Exim filters run in subprocesses under the user's uid. It is
3914 possible for a "deliver" command or an alias in a "personal" command to
3915 provoke an address rewrite. If logging of address rewriting is configured,
3916 this fails because the process is not running as root or exim. There may be
3917 a better way of dealing with this, but for the moment (because 4.30 needs
3918 to be released), I have disabled address rewrite logging when running a
3919 filter in a non-root, non-exim process.
3925 1. The buildconfig auxiliary program wasn't quoting the value set for
3926 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3927 not defined. This bug was masked in 4.22 by the effect that was fixed in
3930 2. Some messages that were rejected after a message id was allocated were
3931 shown as "incomplete" by exigrep. It no longer does this for messages that
3932 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3934 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3935 have allow_domain_literals set, the ID did not get logged in the <= line.
3936 Domain literals are now always recognized in Message-ID: header lines.
3938 4. The first argument for a ${extract expansion item is the key name or field
3939 number. Leading and trailing spaces in this item were not being ignored,
3940 causing some misleading effects.
3942 5. When deliver_drop_privilege was set, single queue runner processes started
3943 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3944 same command in the process it spins off) were not dropping privilege.
3946 6. When the daemon running as "exim" started a queue runner, it always
3947 re-executed Exim in the spun-off process. This is a waste of effort when
3948 deliver_drop_privilege is set. The new process now just calls the
3949 queue-runner function directly.
3955 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3958 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3959 ignored. Though the use of -oP was forcing the writing of a pid file, it
3960 was always written to the default place.
3962 3. If the message "no IP address found for host xxxx" is generated during
3963 incoming verification, it is now followed by identification of the incoming
3964 connection (so you can more easily find what provoked it).
3966 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3968 5. Added some features to "harden" Exim a bit more against certain attacks:
3970 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3971 be put in Local/Makefile. This is like the never_users runtime option,
3972 but it cannot be overridden. The default setting is "root".
3974 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3975 prefix string with which any file named in a -C command line option
3978 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3979 is retained for -C and -D only if the caller of Exim is root. Without
3980 it, the exim user may also use -C and -D and retain privilege.
3982 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3983 command line option is disabled.
3985 6. Macro names set by the -D option must start with an upper case letter, just
3986 like macro names defined in the configuration file.
3988 7. Added "dereference=" facility to LDAP.
3990 8. Two instances of the typo "uknown" in the source files are fixed.
3992 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3993 the Configure-Makefile script screwed up while processing it.
3995 10. Incorporated PCRE 4.4.
3997 11. The SMTP synchronization check was not operating right at the start of an
3998 SMTP session. For example, it could not catch a HELO sent before the client
3999 waited for the greeting. There is now a check for outstanding input at the
4000 point when the greeting is written. Because of the duplex, asynchronous
4001 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4002 way, but not yet received, when the check is performed.
4004 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4005 on TCP/IP sockets, because this apparently causes some broken clients to
4008 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4009 unchanged) from the Cygwin maintainer.
4011 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4012 was supported instead of testing for mbx. Effectively a typo.
4014 15. The spa authenticator server code was not checking that the input it
4015 received was valid base64.
4017 16. The debug output line for the "set" modifier in ACLs was not showing the
4018 name of the variable that was being set.
4020 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4022 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4023 lookup on its own. Something else was needed to trigger the lookup. For
4024 example, a match in host_lookup or the need for a host name in a host list.
4025 Now, if $sender_host_name is referenced and the host name has not yet been
4026 looked up, a lookup is performed. If the lookup fails, the variable remains
4027 empty, and $host_lookup_failed is set to "1".
4029 19. Added "eqi" as a case-independent comparison operator.
4031 20. The saslauthd authentication condition could segfault if neither service
4032 nor realm was specified.
4034 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4035 error message that was logged was misleading, and incoming SMTP
4036 connections were dropped. The message is now more accurate, and temporary
4037 errors are given to SMTP connections.
4039 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4040 (except RSET) with a 421 error, until QUIT is received. However, it was
4041 failing to send a response to QUIT.
4043 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4044 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4045 ACL is now run after the helo_try_verify_hosts code.
4047 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4048 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4049 case-independent, so other case variants are also recognized). Apparently
4050 some systems use these upper case variants.
4052 25. If more than two messages were waiting for the same host, and a transport
4053 filter was specified for the transport, Exim sent two messages over the
4054 same TCP/IP connection, and then failed with "socket operation on non-
4055 socket" when it tried to send the third.
4057 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4059 27. The extern definition of crypt16() in expand.c was not being excluded when
4060 the OS had its own crypt16() function.
4062 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4063 as a preferred synonym for return_size_limit, both as an option and as an
4066 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4068 30. Changed the default configuration ACL to relax the local part checking rule
4069 for addresses that are not in any local domains. For these addresses,
4070 slashes and pipe symbols are allowed within local parts, but the sequence
4071 /../ is explicitly forbidden.
4073 31. SPA server authentication was not clearing the challenge buffer before
4076 32. log_message in a "warn" ACL statement was writing to the reject log as
4077 well as to the main log, which contradicts the documentation and doesn't
4078 seem right (because no rejection is happening). So I have stopped it.
4080 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4081 However, I am unable to do any testing of this.
4083 34. Fixed an infelicity in the appendfile transport. When checking directories
4084 for a mailbox, to see if any needed to be created, it was accidentally
4085 using path names with one or more superfluous leading slashes; tracing
4086 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4088 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4089 contents are added to the log line that is written for every discarded
4090 recipient. (Previously a log_message setting was ignored.)
4092 36. The ${quote: operator now quotes the string if it is empty.
4094 37. The install script runs exim in order to find its version number. If for
4095 some reason other than non-existence or emptiness, which it checks, it
4096 could not run './exim', it was installing it with an empty version number,
4097 i.e. as "exim-". This error state is now caught, and the installation is
4100 38. An argument was missing from the function that creates an error message
4101 when Exim fails to connect to the socket for saslauthd authentication.
4102 This could cause Exim to crash, or give a corrupted message.
4104 39. Added isip, isip4, and isip6 to ${if conditions.
4106 40. The ACL variables $acl_xx are now saved with the message, and can be
4107 accessed later in routers, transports, and filters.
4109 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4110 strings in the file are not string-expanded.
4112 42. If a MAIL command specified a SIZE value that was too large to fit into an
4113 int variable, the check against message_size_limit failed. Such values are
4114 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4115 day this will have to be increased, but I don't think I want to be around
4116 when emails are that large.
4123 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4124 iconv() is not standard in FreeBSD.
4126 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4127 IPv6 enabled. The observed symptom was a segmentation fault on return from
4128 the function os_common_find_running_interfaces() in src/os.c.
4130 3. In the check_special_case() function in daemon.c I had used "errno" as an
4131 argument name, which causes warnings on some systems. This was basically a
4132 typo, since it was named "eno" in the comments!
4134 4. The code that waits for the clock to tick (at a resolution of some fraction
4135 of a second) so as to ensure message-id uniqueness was always waiting for
4136 at least one whole tick, when it could have waited for less. [This is
4137 almost certainly not relevant at current processor speeds, where it is
4138 unlikely to ever wait at all. But we try to future-proof.]
4140 5. The function that sleeps for a time interval that includes fractions of a
4141 second contained a race. It did not block SIGALRM between setting the
4142 timer, and suspending (a couple of lines later). If the interval was short
4143 and the sigsuspend() was delayed until after it had expired, the suspension
4144 never ended. On busy systems this could lead to processes getting stuck for
4147 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4148 process, before it forks any delivery processes. The open lookup caching
4149 mechanism meant that the open file or database connection was passed into
4150 the delivery process. The problem was that delivery processes always tidy
4151 up cached lookup data. This could cause a problem for the next delivery
4152 process started by the queue runner, because the external queue runner
4153 process does not know about the closure. So the next delivery process
4154 still has data in the lookup cache. In the case of a file lookup, there was
4155 no problem because closing a file descriptor in a subprocess doesn't affect
4156 the parent. However, if the lookup was caching a connection to a database,
4157 the connection was closed, and the second delivery process was likely to
4158 see errors such as "PGSQL: query failed: server closed the connection
4159 unexpectedly". The problem has been fixed by closing all cached lookups
4160 in a queue runner before running a delivery process.
4162 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4163 seem to have the "const" qualifier which it has on other OS. I've
4166 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4167 *of the same type* (client or server) with the same public name that an
4168 error should be diagnosed.
4170 9. When Exim looked up a host name for an IP address, but failed to find the
4171 original IP address when looking up the host name (a safety check), it
4172 output the message "<ip address> does not match any IP for NULL", which was
4173 confusing, to say the least. The bug was that the host name should have
4174 appeared instead of "NULL".
4176 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4177 user that is built into the binary, and the -C or -D options is used, root
4178 privilege is dropped before the configuration file is read. In addition,
4179 logging is switched to stderr instead of the normal log files. If the
4180 configuration then re-defines the Exim user, the unprivileged environment
4181 is probably not what is expected, so Exim logs a panic warning message (but
4184 However, if deliver_drop_privilege is set, the unprivileged state may well
4185 be exactly what is intended, so the warning has been cut out in that case,
4186 and Exim is allowed to try to write to its normal log files.
4192 1. smtp_return_error_details was not giving details for temporary sender
4193 or receiver verification errors.
4195 2. Diagnose a configuration error if two authenticators have the same public
4198 3. Exim used not to create the message log file for a message until the first
4199 delivery attempt. This could be confusing when incoming messages were held
4200 for policy or load reasons. The message log file is now created at the time
4201 the message is received, and an initial "Received" line is written to it.
4203 4. The automatically generated man page for command line options had a minor
4204 bug that caused no ill effects; however, a more serious problem was that
4205 the procedure for building the man page automatically didn't always
4206 operate. Consequently, release 4.20 contains an out-of-date version. This
4207 shouldn't happen again.
4209 5. When building Exim with embedded Perl support, the script that builds the
4210 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4211 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4213 6. The freeze_tell option was not being used for messages that were frozen on
4214 arrival, either by an ACL or by local_scan().
4216 7. Added the smtp_incomplete_transaction log selector.
4218 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4219 was accepting AUTH without a new EHLO.
4221 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4222 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4225 10. Logging of TCP/IP connections (when configured) now happens in the main
4226 daemon process instead of the child process, so that the TCP/IP connection
4227 count is more accurate (but it can never be perfect).
4229 11. The use of "drop" in a nested ACL was not being handled correctly in the
4230 outer ACL. Now, if condition failure induced by the nested "drop" causes
4231 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4232 or "require"), the connection is dropped.
4234 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4235 that yield "discard" can now be used with an "accept" or a "discard" verb,
4236 but an error is generated for any others (because I can't see a useful way
4237 to define what should happen).
4239 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4240 are now processed in the same way as lines in the Exim configuration file.
4241 In particular, continuation lines are supported.
4243 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4245 15. Added -ti meaning -t -i.
4247 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4248 domains, and warn by logging if others are found.
4250 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4251 returned from GIFCONF when Exim is trying to find the list of interfaces on
4252 a host. The code in os.c has been modified to copy each ifreq to an aligned
4253 structure in all cases.
4255 Also, in some cases, the returned ifreq's were being copied to a 'struct
4256 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4257 means the last couple of bytes of an IPv6 address could be chopped if the
4258 ifreq contained only a normal sockaddr (14 bytes storage).
4260 18. Named domain lists were not supported in the hosts_treat_as_local option.
4261 An entry such as +xxxx was not recognized, and was treated as a literal
4264 19. Ensure that header lines added by a DATA ACL are included in the reject log
4265 if the ACL subsequently rejects the message.
4267 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4268 MD5 (which is deprecated).
4270 21. When testing a filter file using -bf, Exim was writing a message when it
4271 took the sender from a "From " line in the message, but it was not doing so
4272 when it took $return_path from a Return-Path: header line. It now does.
4274 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4275 with a valid header line field name (a series of printing characters
4276 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4278 23. Changed "disc" in the source to "disk" to conform to the documentation and
4279 the book and for uniformity.
4281 24. Ignore Sendmail's -Ooption=value command line item.
4283 25. When execve() failed while trying to run a command in a pipe transport,
4284 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4285 could be confused with a return value of 69 from the command itself. This
4286 has been changed to 127, the value the shell returns if it is asked to run
4287 a non-existent command. The wording for the related log line suggests a
4288 non-existent command as the problem.
4290 26. If received_header_text expands to an empty string, do not add a Received:
4291 header line to the message. (Well, it adds a token one on the spool, but
4292 marks it "old" so that it doesn't get used or transmitted.)
4294 27. Installed eximstats 1.28 (addition of -nt option).
4296 28. There was no check for failure on the call to getsockname() in the daemon
4297 code. This can fail if there is a shortage of resources on the system, with
4298 ENOMEM, for example. A temporary error is now given on failure.
4300 29. Contrary to the C standard, it seems that in some environments, the
4301 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4302 program. Exim now does this explicitly; it affects the formatting of
4303 timestamps using strftime().
4305 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4306 complaints. I've now initialized all the variables, to avoid this.
4308 32. Header lines added by a system filter were not being "seen" during
4309 transport-time rewrites.
4311 33. The info_callback() function passed to OpenSSL is set up with type void
4312 (*)(SSL *, int, int), as described somewhere. However, when calling the
4313 function (actually a macro) that sets it up, the type void(*)() is
4314 expected. I've put in a cast to prevent warnings from picky compilers.
4316 34. If a DNS black list lookup found a CNAME record, but there were no A
4317 records associated with the domain it pointed at, Exim crashed.
4319 35. If a DNS black list lookup returned more than one A record, Exim ignored
4320 all but the first. It now scans all returned addresses if a particular IP
4321 value is being sought. In this situation, the contents of the
4322 $dnslist_value variable are a list of all the addresses, separated by a
4325 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4326 to accept a host name and all its aliases if the forward lookup for any of
4327 them yielded the IP address of the incoming connection. Now it accepts only
4328 those names whose forward lookup yields the correct IP address. Any other
4329 names are discarded. This closes a loophole whereby a rogue DNS
4330 administrator could create reverse DNS records to break through a
4331 wildcarded host restriction in an ACL.
4333 37. If a user filter or a system filter that ran in a subprocess used any of
4334 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4335 the wrong values were passed to the pipe command ($thisaddress had the
4336 value of $0, $0 had the value of $1, etc). This bug was introduced by
4337 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4339 38. Improved the line breaking for long SMTP error messages from ACLs.
4340 Previously, if there was no break point between 40 and 75 characters, Exim
4341 left the rest of the message alone. Two changes have been made: (a) I've
4342 reduced the minimum length to 35 characters; (b) if it can't find a break
4343 point between 35 and 75 characters, it looks ahead and uses the first one
4344 that it finds. This may give the occasional overlong line, but at least the
4345 remaining text gets split now.
4347 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4348 file descriptors might be low, and that setting 1000 would always raise it.
4349 It turns out that in some environments, the limit is already over 1000 and
4350 that lowering it causes trouble. So now Exim takes care not to decrease it.
4352 40. When delivering a message, the value of $return_path is set to $sender_
4353 address at the start of routing (routers may change the value). By an
4354 oversight, this default was not being set up when an address was tested by
4355 -bt or -bv, which affected the outcome if any router or filter referred to
4358 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4359 log only when "message" or "log_message" are set. However, if one of the
4360 conditions was an address verification, or a call to a nested ACL, the
4361 messages generated by the underlying test were being passed through. This
4362 no longer happens. The underlying message is available in $acl_verify_
4363 message for both "message" and "log_message" expansions, so it can be
4364 passed through if needed.
4366 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4367 new expansion $bh_ to give the encoded byte string without charset
4368 translation. Translation happens only if iconv() is available; HAVE_ICONV
4369 indicates this at build time. HEADERS_CHARSET gives the charset to
4370 translate to; headers_charset can change it in the configuration, and
4371 "headers charset" can change it in an individual filter file.
4373 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4374 that creates RFC 2047 encoded "words" labels them as that charset instead
4375 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4376 expansion operator; (ii) when Exim creates a From: line for a local
4377 message; (iii) when a header line is rewritten to include a "phrase" part.
4379 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4380 buggy, causing it to skip the first lines of messages whose message ID
4381 ended in 'D'. This would not have bitten before Exim release 4.14, because
4382 message IDs were unlikely to end in 'D' before then. The effect was to have
4383 incorrect size information for certain domains.
4385 45. #include "config.h" was missing at the start of the crypt16.c module. This
4386 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4389 46. If there was a timeout during a "random" callout check, Exim treated it as
4390 a failure of the random address, and carried on sending RSET and the real
4391 address. If the delay was just some slowness somewhere, the response to the
4392 original RCPT would be taken as a response to RSET and so on, causing
4393 mayhem of various kinds.
4395 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4396 when I implemented it. It didn't allow for the fact that some option values
4397 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4398 the right test for positive values.
4400 48. Domain names in DNS records are case-independent. Exim always looks them up
4401 in lower case. Some resolvers return domain names in exactly the case they
4402 appear in the zone file, that is, they may contain uppercase letters. Not
4403 all resolvers do this - some return always lower case. Exim was treating a
4404 change of case by a resolver as a change of domain, similar to a widening
4405 of a domain abbreviation. This triggered its re-routing code and so it was
4406 trying to route what was effectively the same domain again. This normally
4407 caused routing to fail (because the router wouldn't handle the domain
4408 twice). Now Exim checks for this case specially, and just changes the
4409 casing of the domain that it ultimately uses when it transmits the message
4412 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4415 50. If a filter generated a file delivery with a non-absolute name (possible if
4416 no home directory exists for the router), the forbid_file option was not
4419 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4420 the existing equality matching.
4422 52. Exim was using ints instead of ino_t variables in some places where it was
4423 dealing with inode numbers.
4425 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4426 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4427 called TMPDIR, and if it finds it is different, it changes its value.
4429 54. The smtp_printf() function is now made available to local_scan() so
4430 additional output lines can be written before returning. There is also an
4431 smtp_fflush() function to enable the detection of a dropped connection.
4432 The variables smtp_input and smtp_batched_input are exported to
4435 55. Changed the default runtime configuration: the message "Unknown user"
4436 has been removed from the ACL, and instead placed on the localuser router,
4437 using the cannot_route_message feature. This means that any verification
4438 failures that generate their own messages won't get overridden. Similarly,
4439 the "Unrouteable address" message that was in the ACL for unverifiable
4440 relay addresses has also been removed.
4442 56. Added hosts_avoid_esmtp to the smtp transport.
4444 57. The exicyclog script was not checking for the esoteric option
4445 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4446 will work only if exicyclog is run under the appropriate euid.
4448 58. Following a discussion on the list, the rules by which Exim recognises line
4449 endings on incoming messages have been changed. The -dropcr and drop_cr
4450 options are now no-ops, retained only for backwards compatibility. The
4451 following line terminators are recognized: LF CRLF CR. However, special
4452 processing applies to CR:
4454 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4455 nor a local message in the state where . is a terminator.
4457 (ii) If a bare CR is encountered in a header line, an extra space is added
4458 after the line terminator so as not to end the header. The reasoning
4459 behind this is that bare CRs in header lines are most likely either
4460 to be mistakes, or people trying to play silly games.
4462 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4463 was being incorrectly given as 18 bytes larger than it should have been.
4464 This is a VOB (very old bug).
4466 60. This may never have affected anything current, but just in case it has:
4467 When the local host is found other than at the start of a list of hosts,
4468 the local host, those with the same MX, and any that follow, are discarded.
4469 When the list in question was part of a longer list of hosts, the following
4470 hosts (not currently being processed) were also being discarded. This no
4471 longer happens. I'm not sure if this situation could ever has previously
4474 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4477 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4478 header. This is something that is recommended in a new Internet Draft, and
4479 is something that is documented as being done by Sendmail. There are two
4480 possible values. For messages generated by the autoreply transport, Exim
4483 Auto-Submitted: auto-replied
4485 whereas for all other generated messages (e.g. bounces) it adds
4487 Auto-Submitted: auto-generated
4489 63. The "personal" condition in filters now includes a test for the
4490 Auto-Submitted: header. If it contains the string "auto-" the message it
4491 not considered personal.
4493 64. Added rcpt_include_affixes as a generic transport option.
4495 65. Added queue_only_override (default true).
4497 66. Added the syslog_duplication option.
4499 67. If what should have been the first header line of a message consisted of
4500 a space followed by a colon, Exim was mis-interpreting it as a header line.
4501 It isn't of course - it is syntactically invalid and should therefore be
4502 treated as the start of the message body. The misbehaviour could have
4503 caused a number of strange effects, including loss of data in subsequent
4504 header lines, and spool format errors.
4506 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4507 client host had authenticated. This control can now be exercised by an ACL
4508 for more flexibility.
4510 69. By default, callouts do not happen when testing with -bh. There is now a
4511 variant, -bhc, which does actually run the callout code, including
4512 consulting and updating the callout cache.
4514 70. Added support for saslauthd authentication, courtesy of Alexander
4517 71. If statvfs() failed on the spool or log directories while checking their
4518 size for availability, Exim confusingly gave the error "space shortage".
4519 Furthermore, in debugging mode it crashed with a floating point exception.
4520 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4521 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4522 serious problem, Exim now writes to the main and panic logs when this
4523 happens, with details of the failure. It then refuses to accept the
4524 incoming message, giving the message "spool directory problem" or "log
4525 directory problem" with a 421 code for SMTP messages.
4527 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4528 0, 1, and 2 exist, because some OS complain for execs without them (see
4529 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4530 descriptors. However, the code omitted to check that the open succeeded,
4531 causing mysterious errors if for some reason the permissions on /dev/null
4532 got screwed. Now Exim writes a message to the main and panic logs, and
4533 bombs out if it can't open /dev/null.
4535 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4536 interact so that it is all more flexible. It is supposed to remain
4537 backwards compatible. Also added extra_local_interfaces.
4539 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4540 to bomb out with an assertion failure - to the client this appears as a
4541 connection drop. This problem occurs in the part of the code that was taken
4542 from the Samba project. Fortunately, the assertion is in a very simple
4543 function, so I have fixed this by reproducing the function inline in the
4544 one place where it is called, and arranging for authentication to fail
4545 instead of killing the process with assert().
4547 75. The SPA client code was not working when the server requested OEM rather
4548 than Unicode encoding.
4550 76. Added code to make require_files with a specific uid setting more usable in
4551 the case where statting the file as root fails - usually a non-root-mounted
4552 NFS file system. When this happens and the failure is EACCES, Exim now
4553 forks a subprocess and does the per-uid checking as the relevant uid.
4555 77. Added process_log_path.
4557 78. If log_file_path was not explicitly set, a setting of check_log_space or
4558 check_log_inodes was ignored.
4560 79. If a space check for the spool or log partitions fails, the incident is now
4561 logged. Of course, in the latter case the data may get lost...
4563 80. Added the %p formatting code to string_format() so that it can be used to
4564 print addresses in debug_print(). Adjusted all the address printing in the
4565 debugging in store.c to use %p rather than %d.
4567 81. There was a concern that a line of code in smtp_in.c could overflow a
4568 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4569 initially expressed, the concern was not well-founded, because trailing
4570 spaces are removed early. However, if the trailing spaces were followed by
4571 a NULL, they did not get removed, so the overflow was possible. Two fixes
4574 (a) I re-wrote the offending code in a cleaner fashion.
4575 (b) If an incoming SMTP command contains a NULL character, it is rejected
4578 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4579 runs initgroups(), so that if the Exim user is in any additional groups,
4580 they will be used during message reception.
4586 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob