1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.411 2006/10/18 08:55:37 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
38 later the small subsequent patch to fix an introduced bug.
40 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
42 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
44 PH/08 An error is now given if message_size_limit is specified negative.
46 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
47 to be given (somewhat) arbitrary names.
49 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
52 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
53 miscellaneous code fixes
55 PH/10 Added the log_reject_target ACL modifier to specify where to log
58 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
59 hostname. This is wrong, because it relates to the incoming message (and
60 probably the interface on which it is arriving) and not to the outgoing
61 callout (which could be using a different interface). This has been
62 changed to use the value of the helo_data option from the smtp transport
63 instead - this is what is used when a message is actually being sent. If
64 there is no remote transport (possible with a router that sets up host
65 addresses), $smtp_active_hostname is used.
67 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
68 tweaks were necessary in order to get it to work (see also 21 below):
69 (a) The code assumed that strncpy() returns a negative number on buffer
70 overflow, which isn't the case. Replaced with Exim's string_format()
72 (b) There were several signed/unsigned issues. I just did the minimum
73 hacking in of casts. There is scope for a larger refactoring.
74 (c) The code used strcasecmp() which is not a standard C function.
75 Replaced with Exim's strcmpic() function.
76 (d) The code set only $1; it now sets $auth1 as well.
77 (e) A simple test gave the error "authentication client didn't specify
78 service in request". It would seem that Dovecot has changed its
79 interface. Fortunately there's a specification; I followed it and
80 changed what the client sends and it appears to be working now.
82 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
85 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
86 address A is aliased to B and C, where B exists and C does not. Without
87 -v the output is "A verified" because verification stops after a
88 successful redirection if more than one address is generated. However,
89 with -v the child addresses are also verified. Exim was outputting "A
90 failed to verify" and then showing the successful verification for C,
91 with its parentage. It now outputs "B failed to verify", showing B's
92 parentage before showing the successful verification of C.
94 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
95 look up a TXT record in a specific list after matching in a combined
98 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
99 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
100 they consult the DNS. I had assumed they would set it the way they
101 wanted; and indeed my experiments on Linux seem to show that in some
102 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
103 To be on the safe side, however, I have now made the interface to
104 host_find_byname() similar to host_find_bydns(), with an argument
105 containing the DNS resolver options. The host_find_byname() function now
106 sets these options at its start, just as host_find_bydns() does. The smtp
107 transport options dns_qualify_single and dns_search_parents are passed to
108 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
109 of host_find_byname() use the default settings of RES_DEFNAMES
110 (qualify_single) but not RES_DNSRCH (search_parents).
112 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
113 spool_read_header() do less string testing, by means of a preliminary
114 switch on the second character of optional "-foo" lines. (This is
115 overdue, caused by the large number of possibilities that now exist.
116 Originally there were few.) While I was there, I also converted the
117 str(n)cmp tests so they don't re-test the leading "-" and the first
118 character, in the hope this might squeeze out yet more improvement.
120 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
121 flag allowing group syntax was set by the header_syntax check but not
122 turned off, possible causing trouble later; (2) The flag was not being
123 set at all for the header_verify test, causing "group"-style headers to
124 be rejected. I have now set it in this case, and also caused header_
125 verify to ignore an empty address taken from a group. While doing this, I
126 came across some other cases where the code for allowing group syntax
127 while scanning a header line wasn't quite right (mostly, not resetting
128 the flag correctly in the right place). These bugs could have caused
129 trouble for malformed header lines. I hope it is now all correct.
131 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
132 with the "reply" argument non-NULL. The code, however (which originally
133 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
134 but it didn't always do it. This confused somebody who was copying the
135 code for some other use. I have removed all the tests.
137 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
138 feature that was used to support insecure browsers during the U.S. crypto
139 embargo. It requires special client support, and Exim is probably the
140 only MTA that supported it -- and would never use it because real RSA is
141 always available. This code has been removed, because it had the bad
142 effect of slowing Exim down by computing (never used) parameters for the
143 RSA_EXPORT functionality.
145 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
146 authenticator to fail if there's a tab character in the incoming data
147 (there should never be unless someone is messing about, as it's supposed
148 to be base64-encoded). Also added, on Timo's advice, the "secured" option
149 if the connection is using TLS or if the remote IP is the same as the
150 local IP, and the "valid-client-cert option" if a client certificate has
153 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
154 authenticators. This can be used for authorization after authentication
155 succeeds. (In the case of plaintext, it servers for both authentication
158 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
159 if any retry times were supplied.
165 SC/01 Use a glob alias rather than an array ref in eximstats generated
166 parser. This improves both readability and performance.
168 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
169 Don't display local sender or destination tables in eximstats unless
170 there is data to show.
171 Added average volumes into the eximstats top table text output.
173 SC/03 Collect data on the number of addresses (recipients) as well
174 as the number of messages in eximstats.
176 TF/01 Correct an error in the documentation for the redirect router. Exim
177 does (usually) call initgroups() when daemonizing.
179 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
180 with consistent privilege compared to when running as a daemon.
182 TF/03 Note in the spec that $authenticated_id is not set for local
183 submissions from trusted users.
185 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
186 Thanks to Dean Brooks <dean@iglou.com> for the patch.
188 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
189 by adding some example configuration directives to the default
190 configuration file. A little bit of work is required to uncomment the
191 directives and define how usernames and passwords are checked, but
192 there is now a framework to start from.
194 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
195 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
196 without this. I don't know how relevant this is to other LDAP libraries.
198 PH/02 Add the verb name to the "unknown ACL verb" error.
200 PH/03 Magnus Holmgren's patch for filter_prepend_home.
202 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
204 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
205 directory not expanded when it should be if an expanded home directory
206 was set for the address (which is overridden by the transport).
208 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
211 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
212 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
213 because it is too late at that time, and has no effect.
215 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
216 security issue with \' (bugzilla #107). I could not use the
217 PQescapeStringConn() function, because it needs a PGconn value as one of
220 PH/08 When testing addresses using -bt, indicate those final addresses that
221 are duplicates that would not cause an additional delivery. At least one
222 person was confused, thinking that -bt output corresponded to deliveries.
223 (Suppressing duplicates isn't a good idea as you lose the information
224 about possibly different redirections that led to the duplicates.)
226 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
227 systems where poll() doesn't work, in particular OS X.
229 PH/10 Added more information to debugging output for retry time not reached.
231 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
232 operations in malware.c.
234 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
237 PH/13 If write_rejectlog was set false when logging was sent to syslog with
238 syslog_duplication set false, log lines that would normally be written
239 both the the main log and to the reject log were not written to syslog at
242 PH/14 In the default configuration, change the use of "message" in ACL warn
243 statements to "add_header".
245 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
246 not followed by a command (e.g. "seen endif").
248 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
249 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
252 PH/17 Added extra conditions to the default value of delay_warning_condition
256 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
257 { match{$h_precedence:}{(?i)bulk|list|junk} } \
258 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
261 The Auto-Submitted: and various List- headers are standardised, whereas I
262 don't think Precedence: ever was.
264 PH/18 Refactored debugging code in route_finduser() to show more information,
265 in particular, the error code if getpwnam() issues one.
267 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
268 This is apparently needed in addition to the PH/07 change above to avoid
269 any possible encoding problems.
271 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
272 but not after initializing Perl.
274 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
275 output them only if debugging. By default they are written stderr,
276 apparently, which is not desirable.
278 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
281 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
284 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
286 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
287 authenticated or an ident call has been made. Suppress the default
288 values for $authenticated_id and $authenticated_sender (but permit -oMai
289 and -oMas) when testing with -bh.
291 PH/24 Re-jigged the order of the tests in the default configuration so that the
292 tests for valid domains and recipients precede the DNS black list and CSA
293 tests, on the grounds that those ones are more expensive.
295 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
296 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
297 This bug exists in every version of Exim that I still have, right back to
300 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
301 However, an attempt to turn on case-sensitivity in a regex key by
302 including (?-i) didn't work because the subject string was already
303 lowercased, and the effects were non-intuitive. It turns out that a
304 one-line patch can be used to allow (?-i) to work as expected.
310 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
311 other effects) broke the use of negated acl sub-conditions.
313 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
316 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
317 "Deny" causes Exim to reject the incoming connection with a 554 error.
318 Unfortunately, if there is a major crisis, such as a disk failure,
319 tcp-wrappers gives "deny", whereas what one would like would be some
320 kind of temporary error. A kludge has been added to help with this.
321 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
322 554 error is used if errno is still zero or contains ENOENT (which occurs
323 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
326 PH/03 Add -lutil to the default FreeBSD LIBS setting.
328 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
329 errors. Otherwise a message that provokes a temporary error (when other
330 messages do not) can cause a whole host to time out.
332 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
333 addresses were routed directly to files or pipes from a redirect router.
334 File deliveries just didn't batch; pipe deliveries might have suffered
337 PH/06 A failure to get a lock for a hints database would erroneously always say
338 "Failed to get write lock", even when it was really a read lock.
340 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
341 of 0600. This has been changed to use the value of the lockfile_mode
342 option (which defaults to 0600).
344 PH/08 Applied small patch from the Sieve maintainer.
346 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
347 folder from quota calculations, a direct delivery into this folder messed
348 up the contents of the maildirsize file. This was because the regex was
349 used only to exclude .Trash (or whatever) when the size of the mailbox
350 was calculated. There was no check that a delivery was happening into an
351 excluded directory. This bug has been fixed by ignoring all quota
352 processing for deliveries into excluded directories.
354 PH/10 Added the maildirfolder_create_regex option to appendfile.
360 PH/01 The code for finding all the local interface addresses on a FreeBSD
361 system running IPv6 was broken. This may well have applied to all BSD
362 systems, as well as to others that have similar system calls. The broken
363 code found IPv4 interfaces correctly, but gave incorrect values for the
364 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
365 that it would not match correctly against @[] and not recognize the IPv6
368 PH/02 The ipliteral router was not recognizing addresses of the form user@
369 [ipv6:....] because it didn't know about the "ipv6:" prefix.
371 PH/03 Added disable_ipv6.
373 PH/04 Changed $reply_address to use the raw form of the headers instead of the
374 decoded form, because it is most often used to construct To: headers
375 lines in autoreplies, and the decoded form may well be syntactically
376 invalid. However, $reply_address has leading white space removed, and all
377 newlines turned into spaces so that the autoreply transport does not
380 PH/05 If group was specified without a user on a router, and no group or user
381 was specified on a transport, the group from the router was ignored.
383 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
384 for visible compile-time settings that can be used to change these
385 numbers, for those that want even more. Backwards compatibility with old
386 spool files has been maintained. However, going back to a previous Exim
387 release will lost any variables that are in spool files.
389 PH/07 Two small changes when running in the test harness: increase delay when
390 passing a TCP/IP connection to a new process, in case the original
391 process has to generate a bounce, and remove special handling of
392 127.0.0.2 (sic), which is no longer necessary.
394 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
395 be the same on different OS.
397 PH/09 Moved a debug statement in filter processing to avoid a race problem when
400 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
401 whether --show-vars was specified or not
403 JJ/02 exipick: Added support for new ACL variable spool format introduced
406 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
407 syntactically invalid From: or Reply-to: line, and a filter used this to
408 generate an autoreply, and therefore failed to obtain an address for the
409 autoreply, Exim could try to deliver to a non-existent relative file
410 name, causing unrelated and misleading errors. What now happens is that
411 it logs this as a hard delivery error, but does not attempt to create a
414 PH/11 The exinext utility has a -C option for testing purposes, but although
415 the given file was scanned by exinext itself; it wasn't being passed on
416 when Exim was called.
418 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
419 an end-of-file indication when reading a command response.
421 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
422 compiled. In many other places in Exim, IPv6 addresses are always
423 recognized, so I have changed this. It also means that IPv4 domain
424 literals of the form [IPV4:n.n.n.n] are now always recognized.
426 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
427 used if the router is not running as root, for example, when verifying at
428 ACL time, or when using -bh. The debugging output from this situation was
429 non-existent - all you got was a failure to exec. I have made two
432 (a) Failures to set uid/gid, the current directory, or a process leader
433 in a subprocess such as that created by queryprogram now generate
434 suitable debugging ouput when -d is set.
436 (b) The queryprogram router detects when it is not running as root,
437 outputs suitable debugging information if -d is set, and then runs
438 the subprocess without attempting to change uid/gid.
440 PH/15 Minor change to Makefile for building test_host (undocumented testing
443 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
444 additional section of a DNS packet that returns MX or SRV records.
445 Instead, it always explicitly searches for A/AAAA records. This avoids
446 major problems that occur when a DNS server includes only records of one
447 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
448 fixed another bug: if SRV records were looked up and the corresponding
449 address records were *not* found in the additional section, the port
450 values from the SRV records were lost.
452 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
453 using the correct key (the original address) when searching the retry
454 rules in order to find which one to use for generating the retry hint.
456 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
457 adding the default one. Similarly, if it contains a Reply-To: header, the
458 errors_reply_to option, if set, is not used.
460 PH/19 When calculating a retry time, Exim used to measure the "time since
461 failure" by looking at the "first failed" field in the retry record. Now
462 it does not use this if it is later than than the arrival time of the
463 message. Instead it uses the arrival time. This makes for better
464 behaviour in cases where some deliveries succeed, thus re-setting the
465 "first failed" field. An example is a quota failure for a huge message
466 when small messages continue to be delivered. Without this change, the
467 "time since failure" will always be short, possible causing more frequent
468 delivery attempts for the huge message than are intended.
469 [Note: This change was subsequently modified - see PH/04 for 4.62.]
471 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
472 $1, $2, $3) because the numerical variables can be reset during some
473 expansion items (e.g. "match"), thereby losing the authentication data.
475 PH/21 Make -bV show the size of off_t variables so that the test suite can
476 decide whether to run tests for quotas > 2G.
478 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
479 mailbox_size, and mailbox_filecount in the appendfile transport. If a
480 filecount value is greater than 2G or if a quota value is greater than 2G
481 on a system where the size of off_t is not greater than 4, a panic error
484 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
485 never match. The debug and -bh output now contains an explicit error
486 message indicating a malformed IPv4 address or mask.
488 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
489 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
492 PH/25 Do not write to syslog when running in the test harness. The only
493 occasion when this arises is a failure to open the main or panic logs
494 (for which there is an explicit test).
496 PH/26 Added the /no_tell option to "control=freeze".
498 PH/27 If a host name lookup failed very early in a connection, for example, if
499 the IP address matched host_lookup and the reverse lookup yielded a name
500 that did not have a forward lookup, an error message of the form "no IP
501 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
502 could be logged. Now it outputs the IP address instead of "NULL".
504 PH/28 An enabling patch from MH: add new function child_open_exim2() which
505 allows the sender and the authenticated sender to be set when
506 submitting a message from within Exim. Since child_open_exim() is
507 documented for local_scan(), the new function should be too.
509 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
510 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
511 results in an empty string is now treated as unset.
513 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
515 PH/31 Added sender_verify_fail logging option.
517 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
518 needed by RFC 822 but not by RFC 2822 was commented out. I have now
519 tidied the source and removed it altogether.
521 PH/33 When a queue run was abandoned because the load average was too high, a
522 log line was always written; now it is written only if the queue_run log
523 selector is set. In addition, the log line for abandonment now contains
524 information about the queue run such as the pid. This is always present
525 in "start" and "stop" lines but was omitted from the "abandon" line.
527 PH/34 Omit spaces between a header name and the colon in the error message that
528 is given when verify = headers_syntax fails (if there are lots of them,
529 the message gets confusing).
531 PH/35 Change the default for dns_check_names_pattern to allow slashes within
532 names, as there are now some PTR records that contain slashes. This check
533 is only to protect against broken name servers that fall over on strange
534 characters, so the fact that it applies to all lookups doesn't matter.
536 PH/36 Now that the new test suite is complete, we can remove some of the
537 special code in Exim that was needed for the old test suite. For example,
538 sorting DNS records because real resolvers return them in an arbitrary
539 order. The new test suite's fake resolver always returns records in the
542 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
543 bounces) except when queue_only is set, to avoid logging races between
544 the different processes.
546 PH/38 Panic-die if .include specifies a non-absolute path.
548 PH/39 A tweak to the "H" retry rule from its user.
550 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
551 a label. They prevented compilation on older perls.
553 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
554 a warning to be raised on newish perls.
556 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
557 on queue. Changes to match documented behaviour of showing count of
558 messages matching specified criteria.
560 PH/40 Changed the default ident timeout from 30s to 5s.
562 PH/41 Added support for the use of login_cap features, on those BSD systems
563 that have them, for controlling the resources used by pipe deliveries.
565 PH/42 The content-scanning code uses fopen() to create files in which to put
566 message data. Previously it was not paying any attention to the mode of
567 the files. Exim runs with umask(0) because the rest of the code creates
568 files with open(), and sets the required mode explicitly. Thus, these
569 files were ending up world-writeable. This was not a big issue, because,
570 being within the spool directory, they were not world-accessible. I have
571 created a function called modefopen, which takes an additional mode
572 argument. It sets umask(777), creates the file, chmods it to the required
573 mode, then resets the umask. All the relevant calls to fopen() in the
574 content scanning code have been changed to use this function.
576 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
577 to 24 hours. This avoids potential overflow problems when processing G
578 and H retry rules. I suspect nobody ever tinkers with this value.
580 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
582 PH/45 When the plaintext authenticator is running as a client, the server's
583 challenges are checked to ensure they are valid base64 strings. By
584 default, the authentication attempt is cancelled if an invalid string is
585 received. Setting client_ignore_invalid_base64 true ignores these errors.
586 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
587 they are received. Thus, the responses can be made to depend on the
588 challenges. If an invalid string is ignored, an empty string is placed in
591 PH/46 Messages that are created by the autoreply transport now contains a
592 References: header, in accordance with RFCs 2822 and 3834.
594 PH/47 Added authenticated_sender_force to the smtp transport.
596 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
598 PH/49 Installed latest patch from the Sieve maintainer.
600 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
601 was also set, the appendfile transport was unnecessarily scanning a
602 directory of message files (e.g. for maildir delivery) to find the count
603 of files (along with the size), even though it did not need this
604 information. It now does the scan only if it needs to find either the
605 size of the count of files.
607 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
609 PH/52 Two bugs concerned with error handling when the smtp transport is
612 (i) Exim was not creating retry information for temporary errors given
613 for individual recipients after the DATA command when the smtp transport
614 was used in LMTP mode. This meant that they could be retried too
615 frequently, and not timed out correctly.
617 (ii) Exim was setting the flag that allows error details to be returned
618 for LMTP errors on RCPT commands, but not for LMTP errors for individual
619 recipients that were returned after the DATA command.
621 PH/53 This is related to PH/52, but is more general: for any failing address,
622 when detailed error information was permitted to be returned to the
623 sender, but the error was temporary, then after the final timeout, only
624 "retry timeout exceeded" was returned. Now it returns the full error as
625 well as "retry timeout exceeded".
627 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
628 do this, and (what is worse) MTAs that accept it.
630 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
631 will now be deprecated.
633 PH/56 New os.c-cygwin from the Cygwin maintainer.
635 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
636 formats (previously only available in exim formats via -bpr, -bpru,
637 and -bpra. Now also available in native and exiqgrep formats)
639 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
640 with very large, slow to parse queues
642 JJ/08 exipick: added ! as generic prefix to negate any criteria format
644 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
646 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
647 responses to authentication challenges, though it was showing the
648 challenges; (ii) I've removed the CR characters from the debug output for
651 PH/58 Allow for the insertion of a newline as well as a space when a string
652 is turned into more than one encoded-word during RFC 2047 encoding. The
653 Sieve code now uses this.
655 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
656 data_4xx, lost_connection, tls_required.
658 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
659 message was being sent as an SMTP response.
661 PH/61 Add -l and -k options to exicyclog.
663 PH/62 When verifying, if an address was redirected to one new address, so that
664 verification continued, and the new address failed or deferred after
665 having set something in $address_data, the value of $address_data was not
666 passed back to the ACL. This was different to the case when no
667 redirection occurred. The value is now passed back in both cases.
669 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
670 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
671 use login_cap.h, so on its own it isn't the distinguishing feature. The
672 new name refers directly to the setclassresources() function.
674 PH/65 Added configuration files for NetBSD3.
676 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
678 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
679 is preferred over IPv4.
681 PH/68 The bounce_return_message and bounce_return_body options were not being
682 honoured for bounces generated during the reception of non-SMTP messages.
683 In particular, this applied to messages rejected by the ACL. This bug has
684 been fixed. However, if bounce_return_message is true and bounce_return_
685 body is false, the headers that are returned for a non-SMTP message
686 include only those that have been read before the error was detected.
687 (In the case of an ACL rejection, they have all been read.)
689 PH/69 The HTML version of the specification is now built in a directory called
690 spec_html instead of spec.html, because the latter looks like a path with
691 a MIME-type, and this confuses some software.
693 PH/70 Catch two compiler warnings in sieve.c.
695 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
696 function verify_get_ident() calls ip_connect() to connect a socket, but
697 if the "connect()" function timed out, ip_connect() used to close the
698 socket. However, verify_get_ident() also closes the socket later, and in
699 between Exim writes to the log, which may get opened at this point. When
700 the socket was closed in ip_connect(), the log could get the same file
701 descriptor number as the socket. This naturally causes chaos. The fix is
702 not to close the socket in ip_connect(); the socket should be closed by
703 the function that creates it. There was only one place in the code where
704 this was missing, in the iplookup router, which I don't think anybody now
705 uses, but I've fixed it anyway.
707 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
708 well as to direct DNS lookups. Otherwise the handling of names in host
709 lists is inconsistent and therefore confusing.
715 PH/01 Two changes to the default runtime configuration:
717 (1) Move the checks for relay_from_hosts and authenticated clients from
718 after to before the (commented out) DNS black list checks.
720 (2) Add control=submission to the relay_from_hosts and authenticated
721 clients checks, on the grounds that messages accepted by these
722 statements are most likely to be submissions.
724 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
726 (1) Generate an error if the third argument for the ${prvs expansion is
729 (2) Treat a missing third argument of ${prvscheck as if it were an empty
732 (3) Reset the variables that are obtained from the first argument of
733 ${prvscheck and used in the second argument before leaving the code,
734 because their memory is reclaimed, so using them afterwards may do
737 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
738 one (it's much easier than Tom thought :-).
740 (5) Because of (4), we can now allow for the use of $prvscheck_result
741 inside the third argument.
743 PH/03 For some reason, the default setting of PATH when running a command from
744 a pipe transport was just "/usr/bin". I have changed it to
747 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
748 anything to be listed in the output from -bV.
750 PH/05 When a filter generated an autoreply, the entire To: header line was
751 quoted in the delivery log line, like this:
753 => >A.N.Other <ano@some.domain> <original@ddress> ...
755 This has been changed so that it extracts the operative address. There
756 may be more than one such address. If so, they are comma-separated, like
759 => >ano@some.domain,ona@other.domain <original@ddress> ...
761 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
762 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
763 not being looked up in the rDNS to get a host name, Exim was showing the
764 IP address twice in Received: lines, even though the IP addresses were
765 identical. For example:
767 Received: from [1.2.3.4] (helo=[1.2.3.4])
769 However, if the real host name was known, it was omitting the HELO data
770 if it matched the actual IP address. This has been tidied up so that it
771 doesn't show the same IP address twice.
773 PH/07 When both +timestamp and +memory debugging was on, the value given by
774 $tod_xxx expansions could be wrong, because the tod_stamp() function was
775 called by the debug printing, thereby overwriting the timestamp buffer.
776 Debugging no longer uses the tod_stamp() function when +timestamp is set.
778 PH/08 When the original message was included in an autoreply transport, it
779 always said "this is a copy of the message, including all the headers",
780 even if body_only or headers_only was set. It now gives an appropriate
783 PH/09 Applied a patch from the Sieve maintainer which:
785 o fixes some comments
786 o adds the (disabled) notify extension core
787 o adds some debug output for the result of if/elsif tests
788 o points to the current vacation draft in the documentation
789 and documents the missing references header update
793 o fixes a bug in processing the envelope test (when testing
794 multiple envelope elements, the last element determinted the
797 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
798 Electronic Mail") by including:
800 Auto-submitted: auto-generated
802 in the messages that it generates (bounce messages and others, such as
803 warnings). In the case of bounce messages for non-SMTP mesages, there was
804 also a typo: it was using "Auto_submitted" (underscore instead of
805 hyphen). Since every message generated by Exim is necessarily in response
806 to another message, thes have all been changed to:
808 Auto-Submitted: auto-replied
810 in accordance with these statements in the RFC:
812 The auto-replied keyword:
814 - SHOULD be used on messages sent in direct response to another
815 message by an automatic process,
817 - MUST NOT be used on manually-generated messages,
819 - MAY be used on Delivery Status Notifications (DSNs) and Message
820 Disposition Notifications (MDNs),
822 - MUST NOT be used on messages generated by automatic or periodic
823 processes, except for messages which are automatic responses to
826 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
827 to the default Received: header definition.
829 PH/12 Added log selector acl_warn_skipped (default on).
831 PH/13 After a successful wildlsearch lookup, discard the values of numeric
832 variables because (a) they are in the wrong storage pool and (b) even if
833 they were copied, it wouldn't work properly because of the caching.
835 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
836 checking when decoding. Apparently there are clients that generate
837 overlong encoded strings. Why am I not surprised?
839 PH/15 If the first argument of "${if match_address" was not empty, but did not
840 contain an "@" character, Exim crashed. Now it writes a panic log message
841 and treats the condition as false.
843 PH/16 In autoreply, treat an empty string for "once" the same as unset.
845 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
846 extension "envelope-auth". The code is finished and in agreement with
847 other implementations, but there is no documentation so far and in fact,
848 nobody wrote the draft yet. This extension is currently #undef'ed, thus
849 not changing the active code.
851 Print executed "if" and "elsif" statements when debugging is used. This
852 helps a great deal to understand what a filter does.
854 Document more things not specified clearly in RFC3028. I had all this
855 sorted out, when out of a sudden new issues came to my mind. Oops."
857 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
860 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
861 canonical form (as documented). However, after a host name lookup from
862 the IP address, check_host() was doing a simple string comparison with
863 addresses acquired from the DNS when checking that the found name did
864 have the original IP as one of its addresses. Since any found IPv6
865 addresses are likely to be in abbreviated form, the comparison could
866 fail. Luckily, there already exists a function for doing the comparison
867 by converting both addresses to binary, so now that is used instead of
870 PH/20 There was another similar case to PH/19, when a complete host name was
871 given in a host list; looking up its IP address could give an abbreviated
872 form, whereas the current host's name might or might not be abbreviated.
873 The same fix has been applied.
879 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
880 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
883 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
884 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
886 PH/03 Typo: missing ".o" in src/pcre/Makefile.
888 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
889 header line, restrict the check to what is listed in RFCs 2369 and 2929.
890 Also, for "Auto-Submitted", treat anything other than "no" as
891 non-personal, in accordance with RFC 3834. (Previously it treated
892 anything starting "auto-" as non-personal.)
894 TF/01 The control=submission/name=... option had a problem with syntax
895 errors if the name included a slash character. The /name= option
896 now slurps the rest of the string, so it can include any characters
897 but it must come last in the list of options (after /sender_retain
900 PH/05 Some modifications to the interface to the fake nameserver for the new
908 TK/01 Added the "success_on_redirect" address verification option. See
909 NewStuff for rationale and an example.
911 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
913 PH/02 Patch to exigrep to allow it to work on syslog lines.
915 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
916 fread() to skip over the body file's header line, because in Cygwin the
917 header line is locked and is inaccessible.
919 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
920 co-exist for some time) to make it clear that it is the Exim ID that is
921 referenced, not the Message-ID: header line.
923 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
924 string_format() function, because snprintf() does not exist on all
927 PH/06 The use of forbid_filter_existstest now also locks out the use of the
928 ${stat: expansion item.
930 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
931 protocol synchronization error", to keep the pedants happy.
933 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
934 well as for IRIX systems, when gcc is being used. See the host.c source
937 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
939 PH/10 Named domain lists were not working if used in a queue_smtp_domains
942 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
943 transport and to the smtp transport in LMTP mode.
945 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
947 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
948 run a filter in a subprocess. This could lead to confusion in subsequent
949 lookups in the parent process. There should also be a search_tidyup() at
950 the end of the subprocess.
952 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
953 only if the host matched helo_try_verify_hosts, which caused the
954 verification to occur when the EHLO/HELO command was issued. The ACL just
955 tested the remembered result. Now, if a previous verification attempt has
956 not happened, "verify = helo" does it there and then.
958 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
960 TK/03 Fix log output including CR from clamd.
962 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
963 exist provoked a memory error which could cause a segfault.
965 PH/15 Installed PCRE 6.2
967 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
969 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
970 of the problem. Specifically, suggested +O2 rather than +O1 for the
973 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
975 PH/20 If a delivery was routed to a non-standard port by means of an SRV
976 record, the port was not correctly logged when the outgoing_port log
977 selector was set (it logged the transort's default port).
979 PH/21 Added support for host-specific ports to manualroute, queryprogram,
980 fallback_hosts, and "hosts" in the smtp transport.
982 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
983 host errors such as "Connection refused".
985 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
986 authentication with radiusclient 0.4.9:
988 - Error returned from rc_read_config was caught wrongly
989 - Username/password not passed on to radius server due to wrong length.
991 The presumption is that some radiusclient API changes for 4.51/PH/17
992 were not taken care of correctly. The code is still untested by me (my
993 Linux distribution still has 0.3.2 of radiusclient), but it was
994 contributed by a Radius user.
996 PH/24 When doing a callout, the value of $domain wasn't set correctly when
997 expanding the "port" option of the smtp transport.
999 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1000 while reading a MIME header. Thanks to Tom Hughes for a patch.
1002 PH/24 Include config.h inside local_scan.h so that configuration settings are
1005 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1006 that in an ACL for RCPT (for example), you can examine exactly what was
1009 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1010 commands, but it was not correctly comparing the address with the actual
1011 client host address. Thus, it would show the EHLO address in Received:
1012 header lines when this was not necessary.
1014 PH/27 Added the % operator to ${eval:}.
1016 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1017 it should be ignoring failures (because with -C, for example, it has lost
1018 privilege). It wasn't ignoring creation failures other than "already
1021 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1024 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1025 because an input line was too long, either on its own, or by virtue of
1026 too many continuations, the temporary file was not being removed, and the
1027 return code was incorrect.
1029 PH/31 Missing "BOOL" in function definition in filtertest.c.
1031 PH/32 Applied Sieve patches from the maintainer.
1033 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1035 PH/33 Added "verify = not_blind".
1037 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1038 Local/Makefile (with some defaults set). These are used in built scripts
1039 such as exicyclog, but they have never been used in the exim_install
1040 script (though there are many overriding facilities there). I have
1041 arranged that the exim_install script now takes note of these two
1044 PH/35 Installed configuration files for Dragonfly.
1046 PH/36 When a locally submitted message by a trusted user did not contain a
1047 From: header, and the sender address was obtained from -f or from an SMTP
1048 MAIL command, and the trusted user did not use -F to supply a sender
1049 name, $originator_name was incorrectly used when constructing a From:
1050 header. Furthermore, $originator_name was used for submission mode
1051 messages from external hosts without From: headers in a similar way,
1052 which is clearly wrong.
1054 PH/37 Added control=suppress_local_fixups.
1056 PH/38 When log_selector = +received_sender was set, and the addition of the
1057 sender made the log line's construction buffer exactly full, or one byte
1058 less than full, an overflow happened when the terminating "\n" was
1061 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1062 when the result of a list match is failure because a DNS lookup failed.
1064 PH/40 RM_COMMAND is now used in the building process.
1066 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1067 the "build-* directories that it finds.
1069 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1070 address in a domain literal was a prefix of an interface address.
1072 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1073 when verifying a sender address, unless rewrite_headers is false.
1075 PH/44 Wrote a long comment about why errors_to addresses are verified as
1076 recipients, not senders.
1078 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1079 the ratelimit ACL was added.
1081 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1083 PH/46 Added extra information about PostgreSQL errors to the error string.
1085 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1086 avoiding the need to install special zones in a real server. This is
1087 backwards compatible; if it can't find the fake resolver, it drops back.
1088 Thus, both old and new test suites can be run.
1090 TF/02 Added util/ratelimit.pl
1092 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1093 clock is set back in time.
1095 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1096 Candler <B.Candler@pobox.com>.
1098 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1099 OK for addresses that are the result of redirections.
1101 PH/48 A number of further additions for the benefit of the new test suite,
1102 including a fake gethostbyname() that interfaces to the fake DNS resolver
1105 TF/06 The fix for widen_domains has also been applied to qualify_single and
1106 search_parents which are the other dnslookup options that can cause
1109 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1112 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1114 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1115 service" if it encountered an MX record with an empty target hostname.
1116 The message is now "an MX or SRV record indicated no SMTP service".
1118 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1119 if the DNS of the sending site is misconfigured. This is quite a
1120 common situation. This change restores the behaviour of treating a
1121 helo verification defer as a failure.
1123 PH/51 If self=fail was set on a router, the bounce message did not include the
1124 actual error message.
1130 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1132 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1133 command itself ended in error, the underlying message about the transport
1134 filter timeout was being overwritten with the pipe command error. Now the
1135 underlying error message should be appended to the second error message.
1137 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1139 PH/02 Reduce the amount of output that "make" produces by default. Full output
1140 can still be requested.
1142 PH/03 The warning log line about a condition test deferring for a "warn" verb
1143 was being output only once per connection, rather than after each
1144 occurrence (because it was using the same function as for successful
1145 "warn" verbs). This seems wrong, so I have changed it.
1147 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1148 it should not have, which might have caused a crash in the right
1149 circumstances, but probably never did.
1151 PH/04 Installed a modified version of Tony Finch's patch to make submission
1152 mode fix the return path as well as the Sender: header line, and to
1153 add a /name= option so that you can make the user's friendly name appear
1156 TF/03 Added the control = fakedefer ACL modifier.
1158 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1159 Mark Lowes for thorough testing.
1161 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1163 TK/03 Merged latest SRS patch from Miles Wilton.
1165 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1166 with the definition in sysexits.h (which is #included earlier).
1167 Fortunately, Exim does not actually use EX_OK. The code used to try to
1168 preserve the sysexits.h value, by assumimg that macro definitions were
1169 scanned for macro replacements. I have been disabused of this notion,
1170 so now the code just undefines EX_OK before #including unistd.h.
1172 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1173 in the smtp transport. When a block could not be written in a single
1174 write() function, the timeout was being re-applied to each part-write.
1175 This seems wrong - if the receiver was accepting one byte at a time it
1176 would take for ever. The timeout is now adjusted when this happens. It
1177 doesn't have to be particularly precise.
1179 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1180 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1182 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1183 without a domain if the check to <postmaster@domain> fails.
1185 SC/01 Eximstats: added -xls and the ability to specify output files
1186 (patch written by Frank Heydlauf).
1188 SC/02 Eximstats: use FileHandles for outputing results.
1190 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1192 SC/04 Eximstats: fixed display of large numbers with -nvr option
1194 SC/05 Eximstats: fixed merging of reports with empty tables.
1196 SC/06 Eximstats: added the -include_original_destination flag
1198 SC/07 Eximstats: removed tabs and trailing whitespace.
1200 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1202 TK/06 MBOX spool code: Add real "From " MBOX separator line
1203 so the .eml file is really in mbox format (even though
1204 most programs do not really care). Patch from Alex Miller.
1206 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1207 The latter is generated from $received_to and is only set if the
1208 message has one envelope recipient. SA can use these headers,
1209 obviously out-of-the-box. Patch from Alex Miller.
1211 PH/08 The ${def test on a variable was returning false if the variable's
1212 value was "0", contrary to what the specification has always said!
1213 The result should be true unless the variable is empty.
1215 PH/09 The syntax error of a character other than { following "${if
1216 def:variable_name" (after optional whitespace) was not being diagnosed.
1217 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1218 accidental colon was present, for example, could give incorrect results.
1220 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1221 result is used (not including appendfile, where other changes are about
1224 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1225 This involved changing a lot of size variables from int to off_t. It
1226 should work with maildirs and everything.
1228 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1229 spamd dying while we are connected to it.
1231 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1234 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1235 function for BDB 4.3.
1237 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1239 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1240 the sources for PCRE have been re-arranged and more files are now
1243 PH/15 The code I had for printing potentially long long variables in PH/11
1244 above was not the best (it lost precision). The length of off_t variables
1245 is now inspected at build time, and an appropriate printing format (%ld
1246 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1247 to be "long long int" or "long int". This is needed for the internal
1248 formatting function string_vformat().
1250 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1251 the configuration file to be ":syslog", then the script "guesses" where
1252 the logs files are, rather than using the compiled in default. In our
1253 case the guess is not the same as the compiled default, so the script
1254 suddenly stopped working when I started to use syslog. The patch checks
1255 to see if log_file_path is "". If so, it attempts to read it from exim
1256 with no configuration file to get the compiled in version, before it
1257 falls back to the previous guessing code."
1259 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1260 implementing BATV in an Exim configuration. See NewStuff for the gory
1263 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1264 Makefile that are specific to HP-UX.
1266 PH/18 If the "use_postmaster" option was set for a recipient callout together
1267 with the "random" option, the postmaster address was used as the MAIL
1268 FROM address for the random test, but not for the subsequent recipient
1269 test. It is now used for both.
1271 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1272 patch removes a few documentation additions to RFC 3028, because the
1273 latest draft now contains them. It adds the new en;ascii-case comparator
1274 and a new error check for 8bit text in MIME parts. Comparator and
1275 require names are now matched exactly. I enabled the subaddress
1276 extension, but it is not well tested yet (read: it works for me)."
1278 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1279 rework some of the code of TK/09 above to avoid the hardwired use of
1280 "%lld" and "long long". Replaced the call to snprintf() with a call to
1283 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1284 records point to non-existent hosts", "retry timeout exceeded", and
1285 "retry time not reached for any host after a long failure period".
1287 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1288 experimental DomainKeys support:
1290 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1291 (2) On an error such as an illegally used "control", the wrong name for
1292 the control was given.
1294 These problems did NOT occur unless DomainKeys support was compiled.
1296 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1298 PH/24 Added ${if match_ip condition.
1300 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1301 databases so that it will be absolutely obvious if a crash occurs in the
1302 DB library. This is a regular occurrence (often caused by mis-matched
1305 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1306 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1307 picked up on a user's system that detects such things. There doesn't seem
1308 to be a gcc warning option for this - only an attribute that has to be
1309 put on the function's prototype. It seems that in Fedora Core 4 they have
1310 set this on a number of new functions. No doubt there will be more in due
1313 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1314 specify a transport. However, if an address that was verified by such a
1315 router was the subject of a callout, Exim crashed because it tried to
1316 read the rcpt_include_affixes from the non-existent transport. Now it
1317 just assumes that the setting of that option is false. This bug was
1318 introduced by 4.51/PH/31.
1320 PH/28 Changed -d+all to exclude +memory, because that information is very
1321 rarely of interest, but it makes the output a lot bigger. People tend to
1322 do -d+all out of habit.
1324 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1325 code in os-type was giving problems when libc.so lives in lib64, like on
1328 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1329 aren't the modern standard, and it seems that some systems' include files
1330 don't always have them. Exim was already checking for some of the newer
1331 ones like T_AAAA, and defining it itself. I've added checks for all the
1332 record types that Exim uses.
1334 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1335 not automatically generating a new one, as it is supposed to. This
1336 prevented TLS from working. If the file did exist, but contained invalid
1337 data, a new version was generated, as expected. It was only the case of a
1338 non-existent file that was broken.
1340 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1341 with a change in libdomainkeys > 0.64.
1343 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1344 from DNS. If the selector record carries the flag, it now has
1345 precedence over the domain-wide flag.
1347 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1349 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1350 the use of a port name that isn't defined in /etc/services) occurred, the
1351 message was deferred as in a normal delivery, and thus remained on the
1352 spool, instead of being failed because of the mua_wrapper setting. This
1353 is now fixed, and I tidied up some of the mua_wrapper messages at the
1356 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1357 the messages in a hash of arrays rather than using individual hashes.
1358 This is a bit cleaner and results in dramatic memory savings, albeit
1359 at a slight CPU cost.
1361 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1362 as requested by Marc Sherman.
1364 SC/10 Eximstats: added histograms for user specified patterns as requested
1367 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1369 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1370 fopen() in the content-scanning modules that did not already have it.
1376 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1377 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1379 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1381 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1383 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1384 to test Sieve filters that use "vacation".
1386 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1387 that changes the way the GnuTLS parameters are stored in the cache file.
1388 The new format can be generated externally. For backward compatibility,
1389 if the data in the cache doesn't make sense, Exim assumes it has read an
1390 old-format file, and it generates new data and writes a new file. This
1391 means that you can't go back to an older release without removing the
1394 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1395 work if there are any delivery delays because "one_time" forces the
1396 parent to be marked "delivered", so its unseen clone is never tried
1397 again. For this reason, Exim now forbids the simultaneous setting of
1400 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1401 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1402 entry if you want to know the details. The fix, however, neglected to
1403 consider the case when local delivery batching is involved. The test for
1404 "previously delivered" was not happening when checking to see if an
1405 address could be batched with a previous (undelivered) one; under
1406 certain circumstances this could lead to multiple deliveries to the same
1409 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1410 in its include files, and this causes problems building Exim.
1412 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1413 header_syntax) but Exim was just ignoring anything given after a slash.
1414 In particular, this caused confusion with an attempt to use "verify =
1415 reverse_host_lookup/defer_ok". An error is now given when options are
1416 supplied for verify items that do not have them. (Maybe reverse_host_
1417 lookup should have a defer_ok option, but that's a different point.)
1419 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1420 defined by RFC 821) to 2048, because there were problems with some AUTH
1421 commands, and RFC 1869 says the size should be increased for extended
1422 SMTP commands that take arguments.
1424 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1427 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1428 "unknown" error; now it says that the functionality isn't in the binary.
1430 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1431 an address' error message when a string expansion fails (syntax or
1432 whatever). Otherwise the password may appear in the log. Following change
1433 PH/42 below, there is no longer a chance of it appearing in a bounce
1436 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1438 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1439 of its addresses was ever tried. (Bugzilla bug #2.)
1441 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1442 the result incorrectly in the debug output. (It correctly added a newline
1443 to what was transported.)
1445 TF/01 Added $received_time.
1447 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1448 commented out examples of how to interface to a virus scanner and to
1449 SpamAssassin. Also added commented examples of av_scanner and
1450 spamd_address settings.
1452 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1453 and controls are allowed in which ACLs. There were a couple of minor
1454 errors. Some of the entries in the conditions table (which is a table of
1455 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1456 negation of where the condition IS allowed.
1458 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1460 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1461 header file does not have a version number, so I've had to invent a new
1462 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1463 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1464 radiusclient), but it was contributed by a Radius user.
1466 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1467 files or directories, for OpenSSL.
1469 PH/19 When an Exim process that is running as root has to create an Exim log
1470 file, it does so in a subprocess that runs as exim:exim so as to get the
1471 ownership right at creation (otherwise, other Exim processes might see
1472 the file with the wrong ownership). There was no test for failure of this
1473 fork() call, which would lead to the process getting stuck as it waited
1474 for a non-existent subprocess. Forks do occasionally fail when resources
1475 run out. I reviewed all the other calls to fork(); they all seem to check
1478 PH/20 When checking for unexpected SMTP input at connect time (before writing
1479 the banner), Exim was not dealing correctly with a non-positive return
1480 from the read() function. If the client had disconnected by this time,
1481 the result was a log entry for a synchronization error with an empty
1482 string after "input=" when read() returned zero. If read() returned -1
1483 (an event I could not check), uninitialized data bytes were printed.
1484 There were reports of junk text (parts of files, etc) appearing after
1487 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1489 PH/22 Added support for macro redefinition, and (re)definition in between
1490 driver and ACL definitions.
1492 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1493 forgetting to use the resulting value; it was using the unexpanded value.
1495 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1496 hadn't been configured. The fix is from Juergen Kreileder, who
1497 understands it better than I do:
1499 "Here's what I see happening with three configured cyrus_sasl
1500 authenticators configured (plain, login, cram-md5):
1502 On startup auth_cyrus_sasl_init() gets called for each of these.
1503 This means three calls to sasl_listmech() without a specified mech_list.
1504 => SASL tests which mechs of all available mechs actually work
1505 => three warnings about OTP not working
1506 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1508 With the patch, sasl_listmech() also gets called three times. But now
1509 SASL's mech_list option is set to the server_mech specified in the the
1510 authenticator. Or in other words, the answer from sasl_listmech()
1511 gets limited to just the mech you're testing for (which is different
1513 => the return list contains just 'plain' or 'login', 'cram-md5' or
1514 nothing depending on the value of ob->server_mech.
1516 I've just tested the patch: Authentication still works fine,
1517 unavailable mechs specified in the exim configuration are still
1518 caught, and the auth.log warnings about OTP are gone."
1520 PH/25 When debugging is enabled, the contents of the command line are added
1521 to the debugging output, even when log_selector=+arguments is not
1524 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1525 answer is "GNU", and only if the return is "GNU/something" is the answer
1528 PH/27 $acl_verify_message is now set immediately after the failure of a
1529 verification in an ACL, and so is available in subsequent modifiers. In
1530 particular, the message can be preserved by coding like this:
1532 warn !verify = sender
1533 set acl_m0 = $acl_verify_message
1535 Previously, $acl_verify_message was set only while expanding "message"
1536 and "log_message" when a very denied access.
1538 PH/28 Modified OS/os.c-Linux with
1540 -#ifndef OS_LOAD_AVERAGE
1541 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1543 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1544 nomenclature these days.)
1546 PH/29 Installed patch from the Sieve maintainer that adds the options
1547 sieve_useraddress and sieve_subaddress to the redirect router.
1549 PH/30 In these circumstances:
1550 . Two addresses routed to the same list of hosts;
1551 . First host does not offer TLS;
1552 . First host accepts first address;
1553 . First host gives temporary error to second address;
1554 . Second host offers TLS and a TLS session is established;
1555 . Second host accepts second address.
1556 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1557 and peerdn, if requested) that were in fact used only for the second
1560 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1561 attention to any local part prefix or suffix that was matched by the
1562 router that accepted the address. It now behaves in the same way as it
1563 does for delivery: the affixes are removed from the local part unless
1564 rcpt_include_affixes is set on the transport.
1566 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1567 timeout during the DATA phase of an incoming message.
1569 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1570 applied a patch sanctioned by the Sieve maintainer.
1572 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1573 the uid or gid is negative. A case of a negative gid caused this to be
1574 noticed. The fix allows for either to be negative.
1576 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1577 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1580 PH/36 The change PH/12 above was broken. Fixed it.
1582 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1583 the grounds that routing the same address twice would always produce the
1584 same answer. This might have been true once, but it is certainly no
1585 longer true now. Routing a child address may depend on the previous
1586 routing that produced that child. Some complicated redirection strategies
1587 went wrong when messages had multiple recipients, and made Exim's
1588 behaviour dependent on the order in which the addresses were given.
1590 I have moved the duplicate checking until after the routing is complete.
1591 Exim scans the addresses that are assigned to local and remote
1592 transports, and removes any duplicates. This means that more work will be
1593 done, as duplicates will always all be routed, but duplicates are
1594 presumably rare, so I don't expect this is of any significance.
1596 For deliveries to pipes, files, and autoreplies, the duplicate checking
1597 still happens during the routing process, since they are not going to be
1600 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1601 It corrects a timeout issue with spamd. This is Ian's comment: "The
1602 background is that sometimes spamd either never reads data from a
1603 connection it has accepted, or it never writes response data. The exiscan
1604 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1605 blindly assumes that writes won't block so it may never time out."
1607 PH/39 Allow G after quota size as well as K and M.
1609 PH/40 The value set for $authenticated_id in an authenticator may not contain
1610 binary zeroes or newlines because the value is written to log lines and
1611 to spool files. There was no check on this. Now the value is run through
1612 the string_printing() function so that such characters are converted to
1613 printable escape sequences.
1615 PH/41 $message_linecount is a new variable that contains the total number of
1616 lines in the message. Compare $body_linecount, which is the count for the
1619 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1620 bounce and delay warning messages, except in certain special cases, which
1623 (a) An SMTP error message from a remote host;
1624 (b) A message specified in a :fail: redirection;
1625 (c) A message specified in a "fail" command in a system filter;
1626 (d) A message specified in a FAIL return from the queryprogram router;
1627 (e) A message specified by the cannot_route_message router option.
1629 In these cases only, Exim does include the error details in bounce and
1630 warning messages. There are also a few cases where bland messages such
1631 as "unrouteable address" or "local delivery error" are given.
1633 PH/43 $value is now also set for the "else" part of a ${run expansion.
1635 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1636 being worked on, but at least Exim now implements the latest version to
1639 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1640 process to complete was treated as a delivery failure, a timeout while
1641 writing the message to the pipe was logged, but erroneously treated as a
1642 successful delivery. Such timeouts include transport filter timeouts. For
1643 consistency with the overall process timeout, these timeouts are now
1644 treated as errors, giving rise to delivery failures by default. However,
1645 there is now a new Boolean option for the pipe transport called
1646 timeout_defer, which, if set TRUE, converts the failures into defers for
1647 both kinds of timeout. A transport filter timeout is now identified in
1650 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1651 systems where "make" and "gmake" are different, calling "gmake" at top
1652 level broke things. I've arranged for the value of $(MAKE) to be passed
1653 from the Makefile to this script so that it can call the same version of
1657 A note about Exim versions 4.44 and 4.50
1658 ----------------------------------------
1660 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1661 changes of various kinds. As a consequence, a big documentation update was
1662 needed. This delayed the release for rather longer than seemed good, especially
1663 in the light of a couple of (minor) security issues. Therefore, the changes
1664 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1665 release. So 4.44 and 4.50 are in effect two different branches that both start
1668 I have left the 4.50 change log unchanged; it contains all the changes since
1669 4.43. The change log for 4.44 is below; many of its items are identical to
1670 those for 4.50. This seems to be the most sensible way to preserve the
1671 historical information.
1677 1. Minor wording change to the doc/README.SIEVE file.
1679 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1680 computation of the current number of files was incorrect.
1682 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1683 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1686 4. Give more explanation in the error message when the command for a transport
1687 filter fails to execute.
1689 5. There are several places where Exim runs a non-Exim command in a
1690 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1691 was being done only for the command run by the queryprogram router. It is
1692 now done for all such subprocesses. The other cases are: ${run, transport
1693 filters, and the commands run by the lmtp and pipe transports.
1695 6. Added CONFIGURE_GROUP build-time option.
1697 7. Some older OS have a limit of 256 on the maximum number of file
1698 descriptors. Exim was using setrlimit() to set 1000 as a large value
1699 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1700 systems. I've change it so that if it can't get 1000, it tries for 256.
1702 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1703 was an oversight, and furthermore, ever since the addition of extra
1704 controls (e.g. 4.43/32), the checks on when to allow different forms of
1705 "control" were broken. There should now be diagnostics for all cases when a
1706 control that does not make sense is encountered.
1708 9. Added the /retain_sender option to "control=submission".
1710 10. $recipients is now available in the predata ACL (oversight).
1712 11. Tidy the search cache before the fork to do a delivery from a message
1713 received from the command line. Otherwise the child will trigger a lookup
1714 failure and thereby defer the delivery if it tries to use (for example) a
1715 cached ldap connection that the parent has called unbind on.
1717 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1718 of $address_data from the recipient verification was clobbered by the
1719 sender verification.
1721 13. The value of address_data from a sender verification is now available in
1722 $sender_address_data in subsequent conditions in the ACL statement.
1724 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1726 15. Added a new option "connect=<time>" to callout options, to set a different
1729 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1730 was its contents. (It was OK if the option was not defined at all.)
1732 17. A "Completed" log line is now written for messages that are removed from
1733 the spool by the -Mrm option.
1735 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1736 information about exactly what failed.
1738 19. Added -dd to debug only the daemon process.
1740 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1741 handles timeouts, both on the server side and network timeouts. Renamed the
1742 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1744 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1745 It is now set to "smtps".
1747 22. $host_address is now set to the target address during the checking of
1748 ignore_target_hosts.
1750 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1751 being passed; this would have caused $sender_host_name to have been used if
1752 matching the list had actually called for a host name (not very likely,
1753 since this list is usually IP addresses). A host name is now passed as
1756 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1757 code that specifies a non-restarting handler (typically sigaction() in
1758 modern systems) in an attempt to fix a rare and obscure crash bug.
1760 25. Narrowed the window for a race in the daemon that could cause it to ignore
1761 SIGCHLD signals. This is not a major problem, because they are used only to
1762 wake it up if nothing else does.
1764 26. A malformed maildirsize file could cause Exim to calculate negative values
1765 for the mailbox size or file count. Odd effects could occur as a result.
1766 The maildirsize information is now recalculated if the size or filecount
1769 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1770 support for a long time. Removed HAVE_SYS_VFS_H.
1772 28. Installed the latest version of exipick from John Jetmore.
1774 29. In an address list, if the pattern was not a regular expression, an empty
1775 subject address (from a bounce message) matched only if the pattern was an
1776 empty string. Non-empty patterns were not even tested. This was the wrong
1777 because it is perfectly reasonable to use an empty address as part of a
1778 database query. An empty address is now tested by patterns that are
1779 lookups. However, all the other forms of pattern expect the subject to
1780 contain a local part and a domain, and therefore, for them, an empty
1781 address still always fails if the pattern is not itself empty.
1783 30. Exim went into a mad DNS loop when attempting to do a callout where the
1784 host was specified on an smtp transport, and looking it up yielded more
1785 than one IP address.
1787 31. Re-factored the code for checking spool and log partition space into a
1788 function that finds that data and another that does the check. The former
1789 is then used to implement four new variables: $spool_space, $log_space,
1790 $spool_inodes, and $log_inodes.
1792 32. The RFC2047 encoding function was originally intended for short strings
1793 such as real names; it was not keeping to the 75-character limit for
1794 encoded words that the RFC imposes. It now respects the limit, and
1795 generates multiple encoded words if necessary. To be on the safe side, I
1796 have increased the buffer size for the ${rfc2047: expansion operator from
1799 33. It is now permitted to omit both strings after an "if" condition; if the
1800 condition is true, the result is "true". As before, when the second string
1801 is omitted, a false condition yields an empty string. This makes it less
1802 cumbersome to write custom ACL and router conditions.
1804 34. Failure to deliver a bounce message always caused it to be frozen, even if
1805 there was an errors_to setting on the router. The errors_to setting is now
1808 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1809 canonical form (fully expanded) before being placed in
1810 $sender_host_address.
1812 36. The table in the code that translates DNS record types into text (T_A to
1813 "A" for instance) was missing entries for NS and CNAME. It is just possible
1814 that this could have caused confusion if both these types were looked up
1815 for the same domain, because the text type is used as part of Exim's
1816 per-process caching. But the chance of anyone hitting this buglet seems
1819 37. The dnsdb lookup has been extended in a number of ways.
1821 (1) There is a new type, "zns", which walks up the domain tree until it
1822 finds some nameserver records. It should be used with care.
1824 (2) There is a new type, "mxh", which is like "mx" except that it returns
1825 just the host names, not the priorities.
1827 (3) It is now possible to give a list of domains (or IP addresses) to be
1828 looked up. The behaviour when one of the lookups defers can be
1829 controlled by a keyword.
1831 (4) It is now possible to specify the separator character for use when
1832 multiple records are returned.
1834 38. The dnslists ACL condition has been extended: it is now possible to supply
1835 a list of IP addresses and/or domains to be looked up in a particular DNS
1838 39. Added log_selector=+queue_time_overall.
1840 40. When running the queue in the test harness, wait just a tad after forking a
1841 delivery process, to get repeatability of debugging output.
1843 41. Include certificate and key file names in error message when GnuTLS fails
1844 to set them up, because the GnuTLS error message doesn't include the name
1845 of the failing file when there is a problem reading it.
1847 42. Allow both -bf and -bF in the same test run.
1849 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1851 44. The "Exiscan patch" is now merged into the mainline Exim source.
1853 45. Sometimes the final signoff response after QUIT could fail to get
1854 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1855 < 0 before doing a fflush(). This bug looks as though it goes back to the
1856 introduction of TLS in release 3.20, but "sometimes" must have been rare
1857 because the tests only now provoked it.
1859 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1860 (this can affect the format of dates).
1862 47. exim_tidydb, when checking for the continued existence of a message for
1863 which it has found a message-specific retry record, was not finding
1864 messages that were in split spool directories. Consequently, it was
1865 deleting retry records that should have stayed in existence.
1867 48. Steve fixed some bugs in eximstats.
1869 49. The SPA authentication driver was not abandoning authentication and moving
1870 on to the next authenticator when an expansion was forced to fail,
1871 contradicting the general specification for all authenticators. Instead it
1872 was generating a temporary error. It now behaves as specified.
1874 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1875 (the order specifies the preference for clients). The order is now AES256,
1876 AES128, 3DES, ARCFOUR128.
1878 51. Small patch to Sieve code - explicitly set From: when generating an
1881 52. Exim crashed if a remote delivery caused a very long error message to be
1882 recorded - for instance if somebody sent an entire SpamAssassin report back
1883 as a large number of 550 error lines. This bug was coincidentally fixed by
1884 increasing the size of one of Exim's internal buffers (big_buffer) that
1885 happened as part of the Exiscan merge. However, to be on the safe side, I
1886 have made the code more robust (and fixed the comments that describe what
1889 53. Now that there can be additional text after "Completed" in log lines (if
1890 the queue_time_overall log selector is set), a one-byte patch to exigrep
1891 was needed to allow it to recognize "Completed" as not the last thing in
1894 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1895 patch that reportedly fixes this has been added. I am not expert enough to
1896 create a test for it. This is what the patch creator wrote:
1898 "I found a little strange behaviour of ldap code when working with
1899 Windows 2003 AD Domain, where users was placed in more than one
1900 Organization Units. When I tried to give exim partial DN, the exit code
1901 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1902 But simultaneously result of request was absolutely normal ldap result,
1903 so I produce this patch..."
1905 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1906 REFERENCE, so I have modified the code to exclude the patch when that macro
1909 55. Some experimental protocols are using DNS PTR records for new purposes. The
1910 keys for these records are domain names, not reversed IP addresses. The
1911 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1912 leaves it alone. Component reversal etc. now happens only for IP addresses.
1915 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1917 57. Double the size of the debug message buffer (to 2048) so that more of very
1918 long debug lines gets shown.
1920 58. The exicyclog utility now does better if the number of log files to keep
1921 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1923 59. Two changes related to the smtp_active_hostname option:
1925 (1) $smtp_active_hostname is now available as a variable.
1926 (2) The default for smtp_banner uses $smtp_active_hostname instead
1927 of $primary_hostname.
1929 60. The host_aton() function is supposed to be passed a string that is known
1930 to be a valid IP address. However, in the case of IPv6 addresses, it was
1931 not checking this. This is a hostage to fortune. Exim now panics and dies
1932 if the condition is not met. A case was found where this could be provoked
1933 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1934 components; fortuitously, this particular loophole had already been fixed
1935 by change 4.50/55 above.
1937 If there are any other similar loopholes, the new check in host_aton()
1938 itself should stop them being exploited. The report I received stated that
1939 data on the command line could provoke the exploit when Exim was running as
1940 exim, but did not say which command line option was involved. All I could
1941 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1942 running as the user.
1945 61. There was a buffer overflow vulnerability in the SPA authentication code
1946 (which came originally from the Samba project). I have added a test to the
1947 spa_base64_to_bits() function which I hope fixes it.
1950 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1951 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1953 63. The daemon start-up calls getloadavg() while still root for those OS that
1954 need the first call to be done as root, but it missed one case: when
1955 deliver_queue_load_max is set with deliver_drop_privilege. This is
1956 necessary for the benefit of the queue runner, because there is no re-exec
1957 when deliver_drop_privilege is set.
1959 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1960 This has been fixed.
1962 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1963 was in use, was not putting the data itself into the right store pool;
1964 consequently, it could be overwritten for a subsequent message in the same
1965 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1968 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1970 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1971 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1972 the function were treating the return as a boolean value, which happened to
1973 work because 0=false and not-0=true, but is not correct code.
1975 68. The host_aton() function was not handling scoped IPv6 addresses (those
1976 with, for example, "%eth0" on the end) correctly.
1978 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1979 negated items (that is, ~something) in unsigned ints. Some compilers
1980 apparently mutter when there is no cast.
1982 70. If an address verification called from an ACL failed, and did not produce a
1983 user-specific message (i.e. there was only a "system" message), nothing was
1984 put in $acl_verify_message. In this situation, it now puts the system
1987 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1988 session; change 4.31/43 added the unwanted input to the log line - except
1989 that it did not do this in the start of session case. It now does.
1991 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1992 This is wrong and can cause the other end to generate a synchronization
1993 error if it is another Exim or anything else that does the synchronization
1994 check. A QUIT command is no longer sent after a timeout.
1996 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1997 during host lookups.
1999 74. The defer_ok option of callout verification was not working if it was used
2000 when verifying addresses in header lines, that is, for this case:
2002 verify = header_sender/callout=defer_ok
2004 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2005 those file descriptors could be used for SMTP connections. If anything
2006 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2007 could be sent to the SMTP client, causing chaos. The daemon now opens
2008 stdin, stdout, and stderr to /dev/null when it puts itself into the
2011 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2012 log by default. The user can override this with suitable Perl magic.
2014 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2015 the log message when discard triggers, was not working for the DATA ACL or
2016 for the non-SMTP ACL.
2018 78. Error message wording change in sieve.c.
2020 79. If smtp_accept_max_per_host was set, the number of connections could be
2021 restricted to fewer than expected, because the daemon was trying to set up
2022 a new connection before checking whether the processes handling previous
2023 connections had finished. The check for completed processes is now done
2024 earlier. On busy systems, this bug wouldn't be noticed because something
2025 else would have woken the daemon, and it would have reaped the completed
2028 80. If a message was submitted locally by a user whose login name contained one
2029 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2030 It caused a spool format error. I have fixed the spool reading code. A
2031 related problem was that the "from" clause in the Received: line became
2032 illegal because of the space(s). It is now covered by ${quote_local_part.
2034 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2037 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2039 83. Patch from Sieve maintainer to fix unterminated string problem in
2040 "vacation" handling.
2042 84. Some minor changes to the Linux configuration files to help with other
2043 OS variants using glibc.
2045 85. One more patch for Sieve to update vacation handling to latest spec.
2048 ----------------------------------------------------
2049 See the note above about the 4.44 and 4.50 releases.
2050 ----------------------------------------------------
2056 1. Change 4.43/35 introduced a bug that caused file counts to be
2057 incorrectly computed when quota_filecount was set in an appendfile
2060 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2061 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2064 3. Give more explanation in the error message when the command for a transport
2065 filter fails to execute.
2067 4. There are several places where Exim runs a non-Exim command in a
2068 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2069 was being done only for the command run by the queryprogram router. It is
2070 now done for all such subprocesses. The other cases are: ${run, transport
2071 filters, and the commands run by the lmtp and pipe transports.
2073 5. Some older OS have a limit of 256 on the maximum number of file
2074 descriptors. Exim was using setrlimit() to set 1000 as a large value
2075 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2076 systems. I've change it so that if it can't get 1000, it tries for 256.
2078 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2079 was an oversight, and furthermore, ever since the addition of extra
2080 controls (e.g. 4.43/32), the checks on when to allow different forms of
2081 "control" were broken. There should now be diagnostics for all cases when a
2082 control that does not make sense is encountered.
2084 7. $recipients is now available in the predata ACL (oversight).
2086 8. Tidy the search cache before the fork to do a delivery from a message
2087 received from the command line. Otherwise the child will trigger a lookup
2088 failure and thereby defer the delivery if it tries to use (for example) a
2089 cached ldap connection that the parent has called unbind on.
2091 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2092 of $address_data from the recipient verification was clobbered by the
2093 sender verification.
2095 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2096 was its contents. (It was OK if the option was not defined at all.)
2098 11. A "Completed" log line is now written for messages that are removed from
2099 the spool by the -Mrm option.
2101 12. $host_address is now set to the target address during the checking of
2102 ignore_target_hosts.
2104 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2105 being passed; this would have caused $sender_host_name to have been used if
2106 matching the list had actually called for a host name (not very likely,
2107 since this list is usually IP addresses). A host name is now passed as
2110 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2111 code that specifies a non-restarting handler (typically sigaction() in
2112 modern systems) in an attempt to fix a rare and obscure crash bug.
2114 15. Narrowed the window for a race in the daemon that could cause it to ignore
2115 SIGCHLD signals. This is not a major problem, because they are used only to
2116 wake it up if nothing else does.
2118 16. A malformed maildirsize file could cause Exim to calculate negative values
2119 for the mailbox size or file count. Odd effects could occur as a result.
2120 The maildirsize information is now recalculated if the size or filecount
2123 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2124 support for a long time. Removed HAVE_SYS_VFS_H.
2126 18. Updated exipick to current release from John Jetmore.
2128 19. Allow an empty sender to be matched against a lookup in an address list.
2129 Previously the only cases considered were a regular expression, or an
2132 20. Exim went into a mad DNS lookup loop when doing a callout where the
2133 host was specified on the transport, if the DNS lookup yielded more than
2136 21. The RFC2047 encoding function was originally intended for short strings
2137 such as real names; it was not keeping to the 75-character limit for
2138 encoded words that the RFC imposes. It now respects the limit, and
2139 generates multiple encoded words if necessary. To be on the safe side, I
2140 have increased the buffer size for the ${rfc2047: expansion operator from
2143 22. Failure to deliver a bounce message always caused it to be frozen, even if
2144 there was an errors_to setting on the router. The errors_to setting is now
2147 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2148 canonical form (fully expanded) before being placed in
2149 $sender_host_address.
2151 24. Updated eximstats to version 1.33
2153 25. Include certificate and key file names in error message when GnuTLS fails
2154 to set them up, because the GnuTLS error message doesn't include the name
2155 of the failing file when there is a problem reading it.
2157 26. Expand error message when OpenSSL has problems setting up cert/key files.
2160 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2161 (this can affect the format of dates).
2163 28. exim_tidydb, when checking for the continued existence of a message for
2164 which it has found a message-specific retry record, was not finding
2165 messages that were in split spool directories. Consequently, it was
2166 deleting retry records that should have stayed in existence.
2168 29. eximstats updated to version 1.35
2169 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2170 1.35 - bugfix such that pie charts by volume are generated correctly
2172 30. The SPA authentication driver was not abandoning authentication and moving
2173 on to the next authenticator when an expansion was forced to fail,
2174 contradicting the general specification for all authenticators. Instead it
2175 was generating a temporary error. It now behaves as specified.
2177 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2178 (the order specifies the preference for clients). The order is now AES256,
2179 AES128, 3DES, ARCFOUR128.
2181 31. Small patch to Sieve code - explicitly set From: when generating an
2184 32. Exim crashed if a remote delivery caused a very long error message to be
2185 recorded - for instance if somebody sent an entire SpamAssassin report back
2186 as a large number of 550 error lines. This bug was coincidentally fixed by
2187 increasing the size of one of Exim's internal buffers (big_buffer) that
2188 happened as part of the Exiscan merge. However, to be on the safe side, I
2189 have made the code more robust (and fixed the comments that describe what
2192 33. Some experimental protocols are using DNS PTR records for new purposes. The
2193 keys for these records are domain names, not reversed IP addresses. The
2194 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2195 leaves it alone. Component reversal etc. now happens only for IP addresses.
2198 34. The host_aton() function is supposed to be passed a string that is known
2199 to be a valid IP address. However, in the case of IPv6 addresses, it was
2200 not checking this. This is a hostage to fortune. Exim now panics and dies
2201 if the condition is not met. A case was found where this could be provoked
2202 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2203 components; fortuitously, this particular loophole had already been fixed
2204 by change 4.50/55 or 4.44/33 above.
2206 If there are any other similar loopholes, the new check in host_aton()
2207 itself should stop them being exploited. The report I received stated that
2208 data on the command line could provoke the exploit when Exim was running as
2209 exim, but did not say which command line option was involved. All I could
2210 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2211 running as the user.
2214 35. There was a buffer overflow vulnerability in the SPA authentication code
2215 (which came originally from the Samba project). I have added a test to the
2216 spa_base64_to_bits() function which I hope fixes it.
2219 36. The daemon start-up calls getloadavg() while still root for those OS that
2220 need the first call to be done as root, but it missed one case: when
2221 deliver_queue_load_max is set with deliver_drop_privilege. This is
2222 necessary for the benefit of the queue runner, because there is no re-exec
2223 when deliver_drop_privilege is set.
2225 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2226 was in use, was not putting the data itself into the right store pool;
2227 consequently, it could be overwritten for a subsequent message in the same
2228 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2231 38. Sometimes the final signoff response after QUIT could fail to get
2232 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2233 < 0 before doing a fflush(). This bug looks as though it goes back to the
2234 introduction of TLS in release 3.20, but "sometimes" must have been rare
2235 because the tests only now provoked it.
2241 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2242 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2243 Now it returns an integer. A number of calls were still expecting a T/F
2244 return. Fortuitously, in all cases, the tests worked in OK situations,
2245 which is the norm. However, things would have gone wrong on any write
2246 failures on the smtp file descriptor. This function is used when sending
2247 messages over SMTP and also when doing verify callouts.
2249 2. When Exim is called to do synchronous delivery of a locally submitted
2250 message (the -odf or -odi options), it no longer closes stderr before doing
2253 3. Implemented the mua_wrapper option.
2255 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2257 5. Implemented the functions header_remove(), header_testname(),
2258 header_add_at_position(), and receive_remove_recipient(), and exported them
2261 6. If an ACL "warn" statement specified the addition of headers, Exim already
2262 inserted X-ACL-Warn: at the start if there was no header name. However, it
2263 was not making this test for the second and subsequent header lines if
2264 there were newlines in the string. This meant that an invalid header could
2265 be inserted if Exim was badly configured.
2267 7. Allow an ACL "warn" statement to add header lines at the start or after all
2268 the Received: headers, as well as at the end.
2270 8. Added the rcpt_4xx retry error code.
2272 9. Added postmaster_mailfrom=xxx to callout verification option.
2274 10. Added mailfrom=xxxx to the callout verification option, for verify=
2277 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2278 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2280 12. Inserted some casts to stop certain compilers warning when using pointer
2281 differences as field lengths or precisions in printf-type calls (mostly
2282 affecting debugging statements).
2284 13. Added optional readline() support for -be (dynamically loaded).
2286 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2287 same clock tick as a message's arrival, so that its received time was the
2288 same as the "first fail" time on the retry record, and that message
2289 remained on the queue past the ultimate address timeout, every queue runner
2290 would try a delivery (because it was past the ultimate address timeout) but
2291 after another failure, the ultimate address timeout, which should have then
2292 bounced the address, did not kick in. This was a "< instead of <=" error;
2293 in most cases the first failure would have been in the next clock tick
2294 after the received time, and all would be well.
2296 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2297 being recognized when the domain list was tested by the match_domain
2298 condition in an expansion string.
2300 16. Added the ${str2b64: operator.
2302 17. Exim was always calling setrlimit() to set a large limit for the number of
2303 processes, without checking whether the existing limit was already
2304 adequate. (It did check for the limit on file descriptors.) Furthermore,
2305 errors from getrlimit() and setrlimit() were being ignored. Now they are
2306 logged to the main and panic logs, but Exim does carry on, to try to do its
2307 job under whatever limits there are.
2309 18. Imported PCRE 5.0.
2311 19. Trivial typo in log message " temporarily refused connection" (the leading
2314 20. If the log selector return_path_on_delivery was set and an address was
2315 redirected to /dev/null, the delivery process crashed because it assumed
2316 that a return path would always be set for a "successful" delivery. In this
2317 case, the whole delivery is bypassed as an optimization, and therefore no
2320 21. Internal re-arrangement: the function for sending a challenge and reading
2321 a response while authentication was assuming a zero-terminated challenge
2322 string. It's now changed to take a pointer and a length, to allow for
2323 binary data in such strings.
2325 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2327 23. Exim was not respecting finduser_retries when seeking the login of the
2328 uid under which it was called; it was always trying 10 times. (The default
2329 setting of finduser_retries is zero.) Also, it was sleeping after the final
2330 failure, which is pointless.
2332 24. Implemented tls_on_connect_ports.
2334 25. Implemented acl_smtp_predata.
2336 26. If the domain in control=submission is set empty, Exim assumes that the
2337 authenticated id is a complete email address when it generates From: or
2338 Sender: header lines.
2340 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2341 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2342 chown and chgrp in /bin and hostname in /usr/bin.
2344 28. Exim was keeping the "process log" file open after each use, just as it
2345 does for the main log. This opens the possibility of it remaining open for
2346 long periods when the USR1 signal hits a daemon. Occasional processlog
2347 errors were reported, that could have been caused by this. Anyway, it seems
2348 much more sensible not to leave this file open at all, so that is what now
2351 29. The long-running daemon process does not normally write to the log once it
2352 has entered its main loop, and it closes the log before doing so. This is
2353 so that log files can straightforwardly be renamed and moved. However,
2354 there are a couple of unusual error situations where the daemon does write
2355 log entries, and I had neglected to close the log afterwards.
2357 30. The text of an SMTP error response that was received during a remote
2358 delivery was being truncated at 512 bytes. This is too short for some of
2359 the long messages that one sometimes sees. I've increased the limit to
2362 31. It is now possible to make retry rules that apply only when a message has a
2363 specific sender, in particular, an empty sender.
2365 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2366 it possible to be selective about when SMTP synchronization is enforced.
2368 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2370 32. Implemented hosts_connection_nolog.
2372 33. Added an ACL for QUIT.
2374 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2377 35. Added mailbox_size and mailbox_filecount to appendfile.
2379 36. Added control = no_multiline_responses to ACLs.
2381 37. There was a bug in the logic of the code that waits for the clock to tick
2382 in the case where the clock went backwards by a substantial amount such
2383 that the microsecond fraction of "now" was more than the microsecond
2384 fraction of "then" (but the whole seconds number was less).
2386 38. Added support for the libradius Radius client library this is found on
2387 FreeBSD (previously only the radiusclient library was supported).
2393 1. When certain lookups returned multiple values in the form name=value, the
2394 quoting of the values was not always being done properly. Specifically:
2395 (a) If the value started with a double quote, but contained no whitespace,
2397 (b) If the value contained whitespace other than a space character (i.e.
2398 tabs or newlines or carriage returns) it was not quoted.
2399 This fix has been applied to the mysql and pgsql lookups by writing a
2400 separate quoting function and calling it from the lookup code. The fix
2401 should probably also be applied to nisplus, ibase and oracle lookups, but
2402 since I cannot test any of those, I have not disturbed their existing code.
2404 2. A hit in the callout cache for a specific address caused a log line with no
2405 reason for rejecting RCPT. Now it says "Previous (cached) callout
2406 verification failure".
2408 3. There was an off-by-one bug in the queryprogram router. An over-long
2409 return line was truncated at 256 instead of 255 characters, thereby
2410 overflowing its buffer with the terminating zero. As well as fixing this, I
2411 have increased the buffer size to 1024 (and made a note to document this).
2413 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2414 when Exim is waiting for an SMTP response from a remote server, Exim
2415 restarts its select() call on the socket, thereby resetting its timeout.
2416 This is not a problem when such interrupts are rare. Somebody set up a cron
2417 job to run exiwhat every 2 minutes, which is less than the normal select()
2418 timeout (5 or 10 minutes). This meant that the select() timeout never
2419 kicked in because it was always reset. I have fixed this by comparing the
2420 time when an interrupt arrives with the time at the start of the first call
2421 to select(). If more time than the timeout has elapsed, the interrupt is
2422 treated as a timeout.
2424 5. Some internal re-factoring in preparation for the addition of Sieve
2425 extensions (by MH). In particular, the "personal" test is moved to a
2426 separate function, and given an option for scanning Cc: and Bcc: (which is
2427 not set for Exim filters).
2429 6. When Exim created an email address using the login of the caller as the
2430 local part (e.g. when creating a From: or Sender: header line), it was not
2431 quoting the local part when it contained special characters such as @.
2433 7. Installed new OpenBSD configuration files.
2435 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2436 try to make them clearer.
2438 9. Callout options, other than the timeout value, were being ignored when
2439 verifying sender addresses in header lines. For example, when using
2441 verify = header_sender/callout=no_cache
2443 the cache was (incorrectly) being used.
2445 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2446 only the Cygwin environment.
2448 11. When return_path_on_delivery was set as a log selector, if different remote
2449 addresses in the same message used different return paths and parallel
2450 remote delivery occurred, the wrong values would sometimes be logged.
2451 (Whenever a remote delivery process finished, the return path value from
2452 the most recently started remote delivery process was logged.)
2454 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2455 lines when AUTH and/or TLS are in use. This is the "received protocol"
2456 field. Exim used to use "asmtp" for authenticated SMTP, without any
2457 indication (in the protocol name) for TLS use. Now it follows the RFC and
2458 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2459 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2460 names appear in log lines as well as in Received: header lines.
2462 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2463 extensions, and comparison tests, and to fix some bugs.
2465 14. Changes to the "personal" filter test:
2467 (1) The test was buggy in that it was just doing the equivalent of
2468 "contains" tests on header lines. For example, if a user's address was
2469 anne@some.where, the "personal" test would incorrectly be true for
2471 To: susanne@some.where
2473 This test is now done by extracting each address from the header in turn,
2474 and checking the entire address. Other tests that are part of "personal"
2475 are now done using regular expressions (for example, to check local parts
2476 of addresses in From: header lines).
2478 (2) The list of non-personal local parts in From: addresses has been
2479 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2480 taken from the Sieve specification recommendations.
2482 (3) If the message contains any header line starting with "List-" it is
2483 treated as non-personal.
2485 (4) The test for "circular" in the Subject: header line has been removed
2486 because it now seems ill-conceived.
2488 15. Minor typos in src/EDITME comments corrected.
2490 16. Installed latest exipick from John Jetmore.
2492 17. If headers_add on a router specified a text string that was too long for
2493 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2494 of string_sprintf() is now avoided.
2496 18. $message_body_size was not set (it was always zero) when running the DATA
2497 ACL and the local_scan() function.
2499 19. For the "mail" command in an Exim filter, no default was being set for
2500 the once_repeat time, causing a random time value to be used if "once" was
2501 specified. (If the value happened to be <= 0, no repeat happened.) The
2502 default is now 0s, meaning "never repeat". The "vacation" command was OK
2503 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2504 (I found it when inspecting the code).
2506 20. There is now an overall timeout for performing a callout verification. It
2507 defaults to 4 times the callout timeout, which applies to individual SMTP
2508 commands during the callout. The overall timeout applies when there is more
2509 than one host that can be tried. The timeout is checked before trying the
2510 next host. This prevents very long delays if there are a large number of
2511 hosts and all are timing out (e.g. when the network connections are timing
2512 out). The value of the overall timeout can be changed by specifying an
2513 additional sub-option for "callout", called "maxwait". For example:
2515 verify = sender/callout=5s,maxwait=20s
2517 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2518 to the end before writing, but this should make it even safer).
2520 22. Exim was forgetting that it had advertised PIPELINING for the second and
2521 subsequent messages on an SMTP connection. It was also not resetting its
2522 memory on STARTTLS and an internal HELO.
2524 23. When Exim logs an SMTP synchronization error within a session, it now
2525 records whether PIPELINING has been advertised or not.
2527 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2528 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2529 rather than long int.
2531 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2533 26. Added the never_mail option to autoreply.
2539 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2540 crash if the getsockname() call failed; this can happen if a connection is
2541 closed very soon after it is established. The problem was simply in the
2542 order in which certain operations were done, causing Exim to try to write
2543 to the SMTP stream before it had set up the file descriptor. The bug has
2544 been fixed by making things happen in the correct order.
2550 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2551 before the connection was closed, thus losing the rejection response.
2553 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2554 some early Solaris releases, but causes trouble in current releases where
2555 socklen_t is defined.
2557 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2560 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2563 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2564 that releasing the top of it at the end releases what was used for sub-
2565 expansions (unless the block got too big). However, discard this block if
2566 the first thing is a variable or header, so that we can use its block when
2567 it is dynamic (useful for very large $message_headers, for example).
2569 6. Lookups now cache *every* query, not just the most recent. A new, separate
2570 store pool is used for this. It can be recovered when all lookup caches are
2571 flushed. Lookups now release memory at the end of their result strings.
2572 This has involved some general refactoring of the lookup sources.
2574 7. Some code has been added to the store_xxx() functions to reduce the amount
2575 of flapping under certain conditions.
2577 8. log_incoming_interface used to affect only the <= reception log lines. Now
2578 it causes the local interface and port to be added to several more SMTP log
2579 lines, for example "SMTP connection from", and rejection lines.
2581 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2583 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2585 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2586 could be overwritten at the end of the current message (or the start of a
2587 new message if it was set in a HELO ACL). The value is now preserved for
2588 the duration of the SMTP connection.
2590 12. If a transport had a headers_rewrite setting, and a matching header line
2591 contained an unqualified address, that address was qualified, even if it
2592 did not match any rewriting rules. The underlying bug was that the values
2593 of the flags that permit the existence of unqualified sender and recipient
2594 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2595 non-local messages, and by -bnq for local messages) were not being
2596 preserved with the message after it was received.
2598 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2599 "next input=" as part of the text comprising the host identity instead of
2600 the correct text. The code was using the same buffer for two different
2601 strings. However, depending on which order the printing function evaluated
2602 its arguments, the bug did not always show up. Under Linux, for example, my
2603 test suite worked just fine.
2605 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2606 doesn't work with some older versions of Perl. It has been changed to "my",
2607 which in any case is probably the better facility to use.
2609 15. A really picky compiler found some instances of statements for creating
2610 error messages that either had too many or two few arguments for the format
2613 16. The size of the buffer for calls to the DNS resolver has been increased
2614 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2615 for addresses that have a lot of PTR records. This alleviates a problem; it
2616 does not fully solve it.
2618 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2619 buffer now truncates the list and logs the incident, which is the same
2620 action as happens when Exim is looking up a host name and its aliases.
2621 Previously in this situation something unpredictable would happen;
2622 sometimes it was "internal error: store_reset failed".
2624 18. If a server dropped the connection unexpectedly when an Exim client was
2625 using GnuTLS and trying to read a response, the client delivery process
2626 crashed while trying to generate an error log message.
2628 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2629 string, for example:
2631 warn message = H1: something\nH2: something
2633 the text was added as a single header line from Exim's point of view
2634 though it ended up OK in the delivered message. However, searching for the
2635 second and subsequent header lines using $h_h2: did not work. This has been
2636 fixed. Similarly, if a system filter added multiple headers in this way,
2637 the routers could not see them.
2639 20. Expanded the error message when iplsearch is called with an invalid key to
2640 suggest using net-iplsearch in a host list.
2642 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2643 ACLs are no longer actually imposed (and a message to that effect is
2646 22. If a "gecos" field in a passwd entry contained escaped characters, in
2647 particular, if it contained a \" sequence, Exim got it wrong when building
2648 a From: or a Sender: header from that name. A second bug also caused
2649 incorrect handling when an unquoted " was present following a character
2650 that needed quoting.
2652 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2653 was not being matched caselessly.
2655 24. Arranged for all hyphens in the exim.8 source to be escaped with
2658 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2659 properly. Recipient callout cache records were still being keyed to include
2660 the sender, even when use_sender was set false. This led to far more
2661 callouts that were necessary. The sender is no longer included in the key
2662 when use_sender is false.
2664 26. Added "control = submission" modifier to ACLs.
2666 27. Added the ${base62d: operator to decode base 62 numbers.
2668 28. dnsdb lookups can now access SRV records.
2670 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2671 the configuration file.
2673 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2674 (-v) mode. This makes the output for a verbose queue run more intelligible.
2676 31. Added a use_postmaster feature to recipient callouts.
2678 32. Added the $body_zerocount variable, containing the number of binary zero
2679 bytes in the message body.
2681 33. The time of last modification of the "new" subdirectory is now used as the
2682 "mailbox time last read" when there is a quota error for a maildir
2685 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2687 35. Added +ignore_unknown as a special item in host lists.
2689 36. Code for decoding IPv6 addresses in host lists is now included, even if
2690 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2691 address was recognized as an IP address, but was then not correctly decoded
2692 into binary, causing unexpected and incorrect effects when compared with
2699 1. Very minor rewording of debugging text in manualroute to say "list of
2700 hosts" instead of "hostlist".
2702 2. If verify=header_syntax was set, and a header line with an unqualified
2703 address (no domain) and a large number of spaces between the end of the
2704 name and the colon was received, the reception process suffered a buffer
2705 overflow, and (when I tested it) crashed. This was caused by some obsolete
2706 code that should have been removed. The fix is to remove it!
2708 3. When running in the test harness, delay a bit after writing a bounce
2709 message to get a bit more predictability in the log output.
2711 4. Added a call to search_tidyup() just before forking a reception process. In
2712 theory, someone could use a lookup in the expansion of smtp_accept_max_
2713 per_host which, without the tidyup, could leave open a database connection.
2715 5. Added the variables $recipient_data and $sender_data which get set from a
2716 lookup success in an ACL "recipients" or "senders" condition, or a router
2717 "senders" option, similar to $domain_data and $local_part_data.
2719 6. Moved the writing of debug_print from before to after the "senders" test
2722 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2723 problems for message scanning, either using a data ACL, or using
2724 local_scan() because the Received: header was not generated till after they
2725 were called (in order to set the time as the time of reception completion).
2726 I have revised the way this works. The header is now generated after the
2727 body is received, but before the ACL or local_scan() are called. After they
2728 are run, the timestamp in the header is updated.
2734 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2735 before starting a queue runner without re-exec. This happened only when
2736 deliver_drop_privilege was set or when the Exim user was set to root. The
2737 effect of the bug was that timeouts during subsequent deliveries caused
2738 crashes instead of being properly handled. The handler is now left at its
2739 default (and expected) setting.
2741 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2742 message, again when deliver_drop_privilege is set or Exim is run as root.
2743 The bug described in (1) was not present in this case, but the tidying up
2744 of the other signals was missing. I have made the two cases consistent.
2746 3. The ignore_target_hosts setting on a manualroute router was being ignored
2747 for hosts that were looked up using the /MX notation.
2749 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2752 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2753 operated on the sender address. After changing the $sender_address to <>
2754 for the sender address verify, Exim was re-instated it as the original
2755 (before rewriting) address, but remembering that it had rewritten it, so it
2756 wasn't rewriting it again. This bug also had the effect of breaking the
2757 sender address verification caching when the sender address was rewritten.
2759 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2760 This has been changed so that if the ip literal address matches
2761 ignore_target_hosts, the router declines.
2763 7. Added expansion conditions match_domain, match_address, and match_local_
2764 part (NOT match_host).
2766 8. The placeholder for the Received: header didn't have a length field set.
2768 9. Added code to Exim itself and to exim_lock to test for a specific race
2769 condition that could lead to file corruption when using MBX delivery. The
2770 issue is with the lockfile that is created in /tmp. If this file is removed
2771 after a process has opened it but before that process has acquired a lock,
2772 there is the potential for a second process to recreate the file and also
2773 acquire a lock. This could lead to two Exim processes writing to the file
2774 at the same time. The added code performs the same test as UW imapd; it
2775 checks after acquiring the lock that its file descriptor still refers to
2776 the same named file.
2778 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2779 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2782 11. Added the smtp_active_hostname option. If used, this will typically be made
2783 to depend on the incoming interface address. Because $interface_address is
2784 not set up until the daemon has forked a reception process, error responses
2785 that can happen earlier (such as "too many connections") no longer contain
2788 12. If an expansion in a condition on a "warn" statement fails because a lookup
2789 defers, the "warn" statement is abandoned, and the next ACL statement is
2790 processed. Previously this caused the whole ACL to be aborted.
2792 13. Added the iplsearch lookup type.
2794 14. Added ident_timeout as a log selector.
2796 15. Added tls_certificate_verified as a log selector.
2798 16. Added a global option tls_require_ciphers (compare the smtp transport
2799 option of the same name). This controls incoming TLS connections.
2801 17. I finally figured out how to make tls_require_ciphers do a similar thing
2802 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2803 before starting the TLS session.
2805 18. Tabs are now shown as \t in -bP output.
2807 19. If the log selector return_path_on_delivery was set, Exim crashed when
2808 bouncing a message because it had too many Received: header lines.
2810 20. If two routers both had headers_remove settings, and the first one included
2811 a superfluous trailing colon, the final name in the first list and the
2812 first name in the second list were incorrectly joined into one item (with a
2813 colon in the middle).
2819 1. Added -C and -D options to the exinext utility, mainly to make it easier
2820 to include in the automated testing, but these could be helpful when
2821 multiple configurations are in use.
2823 2. The exinext utility was not formatting the output nicely when there was
2824 an alternate port involved in the retry record key, nor when there was a
2825 message id as well (for retries that were specific to a specific message
2826 and a specific host). It was also confused by IPv6 addresses, because of
2827 the additional colons they contain. I have fixed the IPv4 problem, and
2828 patched it up to do a reasonable job for IPv6.
2830 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2831 delivery, the log line now contains "pipelined" if PIPELINING was used.
2833 4. An SMTP transport process used to panic and die if the bind() call to set
2834 an explicit outgoing interface failed. This has been changed; it is now
2835 treated in the same way as a connect() failure.
2837 5. A reference to $sender_host_name in the part of a conditional expansion
2838 that was being skipped was still causing a DNS lookup. This no longer
2841 6. The def: expansion condition was not recognizing references to header lines
2842 that used bh_ and bheader_.
2844 7. Added the _cache feature to named lists.
2846 8. The code for checking quota_filecount in the appendfile transport was
2847 allowing one more file than it should have been.
2849 9. For compatibility with Sendmail, the command line option
2857 and sets the incoming protocol and host name (for trusted callers). The
2858 host name and its colon can be omitted when only the protocol is to be set.
2859 Note the Exim already has two private options, -pd and -ps, that refer to
2860 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2861 "s", but I don't think that's a major issue.
2863 10. A number of refactoring changes to the code, none of which should affect
2866 (a) The number of logging options was getting close to filling up the
2867 32-bit word that was used as a bit map. I have split them into two classes:
2868 those that are passed in the argument to log_write(), and those that are
2869 only ever tested independently outside of that function. These are now in
2870 separate 32-bit words, so there is plenty of room for expansion again.
2871 There is no change in the user interface or the logging behaviour.
2873 (b) When building, for example, log lines, the code previously used a
2874 macro that called string_cat() twice, in order to add two strings. This is
2875 not really sufficiently general. Furthermore, there was one instance where
2876 it was actually wrong because one of the argument was used twice, and in
2877 one call a function was used. (As it happened, calling the function twice
2878 did not affect the overall behaviour.) The macro has been replaced by a
2879 function that can join an arbitrary number of extra strings onto a growing
2882 (c) The code for expansion conditions now uses a table and a binary chop
2883 instead of a serial search (which was left over from when there were very
2884 few conditions). Also, it now recognizes conditions like "pam" even when
2885 the relevant support is not compiled in: a suitably worded error message is
2886 given if an attempt is made to use such a condition.
2888 11. Added ${time_interval:xxxxx}.
2890 12. A bug was causing one of the ddress fields not to be passed back correctly
2891 from remote delivery subprocesses. The field in question was not being
2892 subsequently used, so this caused to problems in practice.
2894 13. Added new log selectors queue_time and deliver_time.
2896 14. Might have fixed a bug in maildirsizefile handling that threw up
2897 "unexpected character" debug warnings, and recalculated the data
2898 unnecessarily. In any case, I expanded the warning message to give more
2901 15. Added the message "Restricted characters in address" to the statements in
2902 the default ACL that block characters like @ and % in local parts.
2904 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2905 Three changes have been made:
2907 (a) There was a serious bug; a negative response to MAIL caused the whole
2908 recipient domain to be cached as invalid, thereby blocking all messages
2909 to all local parts at the same domain, from all senders. This bug has
2910 been fixed. The domain is no longer cached after a negative response to
2911 MAIL if the sender used is not empty.
2913 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2916 (c) A new callout option, "use_sender" has been added for people who want
2917 the modified behaviour.
2923 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2926 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2927 indeed breaks things for older releases.
2929 3. Added additional logging to the case where there is a problem reading data
2930 from a filter that is running in a subprocess using a pipe, in order to
2931 try to track down a specific problem.
2933 4. Testing facility fudge: when running in the test harness and attempting
2934 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2935 getting "No route to host". Convert this to a timeout.
2937 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2940 6. Some OS don't have socklen_t but use size_t instead. This affects the
2941 fifth argument of getsockopt() amongst other things. This is now
2942 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2943 can be set for individual OS. I have set it for SunOS5, OSF1, and
2944 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2945 some earlier ones do not.
2947 7. Change 4.30/15 was not doing the test caselessly.
2949 8. The standard form for an IPv6 address literal was being rejected by address
2950 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2951 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2952 this, as well as the form without the "IPv6" on the front (but only when
2953 address literals are enabled, of course).
2955 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2957 10. Exim crashed if a message with an empty sender address specified by -f
2958 encountered a router with an errors_to setting. This could be provoked only
2959 by a command such as
2963 where an empty string was supplied; "<>" did not hit this bug.
2965 11. Installed PCRE release 4.5.
2967 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2968 remained set. It is now erased.
2970 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2971 times from message ids (which are base 36 rather than the normal 62).
2973 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2974 were being counted as actual protocol errors, and logged if the log
2975 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2976 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2977 and DATA following a set of rejected RCPTs do not count as protocol errors.
2978 In other words, Exim assumes they were pipelined, though this may not
2979 actually be the case. Of course, in all cases the client gets an
2980 appropriate error code.
2982 15. If a lookup fails in an ACL condition, a message about the failure may
2983 be available; it is used if testing the ACL cannot continue, because most
2984 such messages specify what the cause of the deferral is. However, some
2985 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2986 that caused an old message to be retained and used if a later statement
2987 caused a defer, replacing the real cause of the deferral.
2989 16. If an IP address had so many PTR records that the DNS lookup buffer
2990 was not large enough to hold them, Exim could crash while trying to process
2991 the truncated data. It now detects and logs this case.
2993 17. Further to 4.21/58, another change has been made: if (and only if) the
2994 first line of a message (the first header line) ends with CRLF, a bare LF
2995 in a subsequent header line has a space inserted after it, so as not to
2996 terminate the header.
2998 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2999 unnecessarily, used atoi() instead of strtol(), and didn't check the
3000 termination when getting file sizes from file names by regex.
3002 19. Completely re-implemented the support for maildirsize files, in the light
3003 of a number of problems with the previous contributed implementation
3004 (4.30/29). In particular:
3006 . If the quota is zero, the maildirsize file is maintained, but no quota is
3009 . If the maildir directory does not exist, it is created before any attempt
3010 to write a maildirsize file.
3012 . The quota value in the file is just a cache; if the quota is changed in
3013 the transport, the new value overrides.
3015 . A regular expression is available for excluding directories from the
3018 20. The autoreply transport checks the characters in options that define the
3019 message's headers; it allows continued headers, but it was checking with
3020 isspace() after an embedded newline instead of explicitly looking for a
3023 21. If all the "regular" hosts to which an address was routed had passed their
3024 expiry times, and had not reached their retry times, the address was
3025 bounced, even if fallback hosts were defined. Now Exim should go on to try
3028 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3029 equivalent code in the SMTP transport. Some hosts send humungous responses
3030 to HELO/EHLO, more than 1024 it seems.
3032 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3033 gives compiler warnings in some environments. I've now done it "properly",
3036 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3037 (because of problems with the built-in one) was declared to return uschar *
3038 instead of char *, causing compiler failure.
3040 25. Fixed a file descriptor leak when processing alias/forward files.
3042 26. Fixed a minor format string issue in dbfn.c.
3044 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3046 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3047 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3050 29. When a sender address is verified, it is cached, to save repeating the test
3051 when there is more than one recipient in a message. However, when the
3052 verification involves a callout, it is possible for different callout
3053 options to be set for different recipients. It is too complicated to keep
3054 track of this in the cache, so now Exim always runs a verification when a
3055 callout is required, relying on the callout cache for the optimization.
3056 The overhead is duplication of the address routing, but this should not be
3059 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3060 to be verified with callout=postmaster, and the main callout worked but the
3061 postmaster check failed, the verification correctly failed. However, if a
3062 subsequent RCPT command asked for sender verification *without* the
3063 postmaster check, incorrect caching caused this verification also to fail,
3066 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3067 it was not caching the DNS options (qualify_single, search_parents) that
3068 were used when the lookup failed. A subsequent lookup with different
3069 options therefore always gave the same answer, though there were cases
3070 where it should not have. (Example: a "domains = !$mx_any" option on a
3071 dnslookup router: the "domains" option is always processed without any
3072 widening, but the router might have qualify_single set.) Now Exim uses the
3073 cached value only when the same options are set.
3075 32. Added John Jetmore's "exipick" utility to the distribution.
3077 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3078 than a timeout (e.g. a certificate is required, and is not provided), an
3079 Exim server now closes the connection immediately. Previously it waited for
3080 the client to close - but if the client is SSL, it seems that they each
3081 wait for each other, leading to a delay before one of them times out.
3083 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3084 maintained 0.8.x compatibility because I don't think many are using it, and
3085 it is clearly obsolete.
3087 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3090 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3091 client certificate was expired. A simple patch fixes this, though I don't
3092 understand the full logic of why the verify callback is called multiple
3095 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3096 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3097 which causes problems with some clients (such as the Certicom SSL Plus
3098 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3099 disables the coutermeasure allowing Eudora to connect."
3101 38. Exim was not checking that a write() to a log file succeeded. This could
3102 lead to Bad Things if a log got too big, in particular if it hit a file
3103 size limit. Exim now panics and dies if it cannot write to a log file, just
3104 as it does if it cannot open a log file.
3106 39. Modified OS/Makefile-Linux so that it now contains
3108 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3110 The two -D definitions ensure that Exim is compiled with large file
3111 support, which makes it possible to handle log files that are bigger than
3114 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3115 instance) a domain was checked against a named list that involved a lookup,
3116 causing $domain_data to be set, then another domain was checked against the
3117 same list, then the first domain was re-checked, the value of $domain_data
3118 after the final check could be wrong. In particular, if the second check
3119 failed, it could be set empty. This bug probably also applied to
3122 41. The strip_trailing_dot option was not being applied to the address given
3123 with the -f command-line option.
3125 42. The code for reading a message's header from the spool was incrementing
3126 $received_count, but never initializing it. This meant that the value was
3127 incorrect (doubled) while delivering a message in the same process in which
3128 it was received. In the most common configuration of Exim, this never
3129 happens - a fresh exec is done - but it can happen when
3130 deliver_drop_privilege is set.
3132 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3133 it now includes up to 150 characters of the unexpected data in the log
3136 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3137 and building data strings. The size of both of these buffers was 10 000
3138 bytes - far larger than anybody would *ever* want, thought I. Needless to
3139 say, somebody hit the limit. I have increased the maximum line length to
3140 20 000 and the maximum data length of concatenated lines to 100 000. I have
3141 also fixed two bugs, because there was no checking on these buffers. Tsk,
3142 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3143 buffer is too small.
3145 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3146 lsearch lookups. Now it does.
3148 46. When parsing a route_list item in a manualroute router, a fixed-length
3149 buffer was used for the list of hosts. I made this 1024 bytes long,
3150 thinking that nobody would ever have a list of hosts that long. Wrong.
3151 Somebody had a whole pile of complicated expansion conditions, and the
3152 string was silently truncated, leading to an expansion error. It turns out
3153 that it is easier to change to an unlimited length (owing to other changes
3154 that have happened since this code was originally written) than to build
3155 structure for giving a limitation error. The length of the item that
3156 expands into the list of hosts is now unlimited.
3158 47. The lsearch lookup could not handle data where the length of text line was
3159 more than 4095 characters. Such lines were truncated, leading to shortened
3160 data being returned. It should now handle lines of any length.
3162 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3163 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3166 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3167 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3168 no longer happens in comment lines. A list of replacements is now placed
3169 at the head of all of the source files, except those whose only change is
3170 to replace PERL_COMMAND in the very first #! line.
3172 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3173 messages on the queue, with a bottom-up merge sort, using code contributed
3174 by Michael Haardt. This should make operations like -bp somewhat faster on
3175 large queues. It won't affect queue runners, except when queue_run_in_order
3178 51. Installed eximstats 1.31 in the distribution.
3180 52. Added support for SRV lookups to the dnslookup router.
3182 53. If an ACL referred to $message_body or $message_body_end, the value was not
3183 reset for any messages that followed in the same SMTP session.
3185 54. The store-handling optimization for building very long strings was not
3186 differentiating between the different store pools. I don't think this
3187 actually made any difference in practice, but I've tidied it.
3189 55. While running the routers to verify a sender address, $sender_address
3190 was still set to the sender address. This is wrong, because when routing to
3191 send a bounce to the sender, it would be empty. Therefore, I have changed
3192 it so that, while verifying a sender address, $sender_address is set to <>.
3193 (There is no change to what happens when verifying a recipient address.)
3195 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3196 target A or AAAA records (if not already returned) without resetting the
3197 qualify_single or search_parents options of the DNS resolver. These are
3198 inappropriate in this case because the targets of MX and SRV records must
3199 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3200 target that, when qualified, matched something in the local domain. These
3201 two options are now turned off when doing these lookups.
3203 57. It seems that at least some releases of Reiserfs (which does not have the
3204 concept of a fixed number of inodes) returns zero and not -1 for the
3205 number of available inodes. This interacted badly with check_spool_inodes,
3206 which assumed that -1 was the "no such thing" setting. What I have done is
3207 to check that the total number of inodes is greater than zero before doing
3208 the test of how many are available.
3210 58. When a "warn" ACL statement has a log_message modifier, the message is
3211 remembered, and not repeated. This is to avoid a lot of repetition when a
3212 message has many recipients that cause the same warning to be written.
3213 Howewer, Exim was preserving the list of already written lines for an
3214 entire SMTP session, which doesn't seem right. The memory is now reset if a
3215 new message is started.
3217 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3218 debugging output unless log_rewrite was also set.
3220 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3221 of (int)(handle) when we know that handle contains (void *)(-1).
3223 61. The Exim daemon panic-logs an error return when it closes the incoming
3224 connection. However "connection reset by peer" seems to be common, and
3225 isn't really an error worthy of noting specially, so that particular error
3228 62. When Exim is trying to find all the local interfaces, it used to panic and
3229 die if the ioctl to get the interface flags failed. However, it seems that
3230 on at least one OS (Solaris 9) it is possible to have an interface that is
3231 included in the list of interfaces, but for which you get a failure error
3232 for this call. This happens when the interface is not "plumbed" into a
3233 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3234 failure of the "get flags" call assumes that the interface is down.
3236 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3237 makes life easier for people who are doing arithmetic on fields extracted
3238 from dates, where you often get leading zeros that should not be
3239 interpreted as octal.
3241 64. Added qualify_domain to the redirect router, to override the global
3244 65. If a pathologically long header line contained very many addresses (the
3245 report of this problem mentioned 10 000) and each of them was rewritten,
3246 Exim could use up a very large amount of memory. (It kept on making new
3247 copies of the header line as it rewrote, and never released the old ones.)
3248 At the expense of a bit more processing, the header rewriting function has
3249 been changed so that it no longer eats memory in this way.
3251 66. The generation of the Received: header has been moved from the time that a
3252 message starts to be received, to the time that it finishes. The timestamp
3253 in the Received: header should now be very close to that of the <= log
3254 line. There are two side-effects of this change:
3256 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3257 logged header lines no longer include the local Received: line, because
3258 it has not yet been created. The same applies to a copy of the message
3259 that is returned to a non-SMTP sender when a message is rejected.
3261 (b) When a filter file is tested using -bf, no additional Received: header
3262 is added to the test message. After some thought, I decided that this
3265 This change does not affect the value of $received_for. It is still set
3266 after address rewriting, but before local_scan() is called.
3268 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3270 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3271 gave an unhelpful panic error message, and a defer error. I have managed to
3272 change this behaviour so that it now rejects any supplied certificate,
3273 which seems right, as the list of acceptable certificates is empty.
3275 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3276 gave an unhelpful defer error. I have not managed to make this reject any
3277 supplied certificates, but the error message it gives is "no certificate
3278 supplied", which is not helpful.
3280 70. exigrep's output now also includes lines that are not associated with any
3281 message, but which match the given pattern. Implemented by a patch from
3282 Martin Sluka, which also tidied up the Perl a bit.
3284 71. Recipient callout verification, like sender verification, was using <> in
3285 the MAIL FROM command. This isn't really the right thing, since the actual
3286 sender may affect whether the remote host accepts the recipient or not. I
3287 have changed it to use the actual sender in the callout; this means that
3288 the cache record is now keyed on a recipient/sender pair, not just the
3289 recipient address. There doesn't seem to be a real danger of callout loops,
3290 since a callout by the remote host to check the sender would use <>.
3291 [SEE ABOVE: changed after hitting problems.]
3293 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3294 temporary errors. However, in the case of such a code being given after
3295 the end of a data transmission (i.e. after ".") Exim was failing to write
3296 a retry record for the message. (Yes, there was some broken host that was
3297 actually sending 8xx at this point.)
3299 73. An unknown lookup type in a host list could cause Exim to panic-die when
3300 the list was checked. (An example that provoked this was putting <; in the
3301 middle of a list instead of at the start.) If this happened during a DATA
3302 ACL check, a -D file could be left lying around. This kind of configuration
3303 error no longer causes Exim to die; instead it causes a defer errror. The
3304 incident is still logged to the main and panic logs.
3306 74. Buglet left over from Exim 3 conversion. The message "too many messages
3307 in one connection" was written to the rejectlog but not the mainlog, except
3308 when address rewriting (yes!) was being logged.
3310 75. Added write_rejectlog option.
3312 76. When a system filter was run not as root (that is, when system_filter_user
3313 was set), the values of the $n variables were not being returned to the
3314 main process; thus, they were not subsequently available in the $sn
3317 77. Added +return_path_on_delivery log selector.
3319 78. A connection timeout was being treated differently from recipients deferred
3320 when testing hosts_max_try with a message that was older than the host's
3321 retry timeout. (The host should not be counted, thus allowing all hosts to
3322 be tried at least once before bouncing.) This may have been the cause of an
3323 occasionally reported bug whereby a message would remain on the queue
3324 longer than the retry timeout, but would be bounced if a delivery was
3325 forced. I say "may" because I never totally pinned down the problem;
3326 setting up timeout/retry tests is difficult. See also the next item.
3328 79. The ultimate address timeout was not being applied to errors that involved
3329 a combination of host plus message (for example, a timeout on a MAIL
3330 command). When an address resolved to a number of possible hosts, and they
3331 were not all tried for each delivery (e.g. because of hosts_max_try), a
3332 message could remain on the queue longer than the retry timeout.
3334 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3337 81. Fixed an obscure SMTP outgoing bug which required at least the following
3338 conditions: (a) there was another message waiting for the same server;
3339 (b) the server returned 5xx to all RCPT commands in the first message so
3340 that the message was not completed; (c) the server dropped the connection
3341 or gave a negative response to the RSET that Exim sends to abort the
3342 transaction. The observed case was a dropped connection after DATA that had
3343 been sent in pipelining mode. That is, the server had advertised PIPELINING
3344 but was not implementing it correctly. The effect of the bug was incorrect
3345 behaviour, such as trying another host, and this could lead to a crash.
3351 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3352 and daemon.c were passed as pointers to ints; they should have been
3353 pointers to socklen_t variables (which are typically unsigned ints).
3355 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3358 3. Fixed a really odd bug that affected only the testing scheme; patching a
3359 certain fixed string in the binary changed the value of another string that
3360 happened to be identical to the end of the original first string.
3362 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3363 name", it returns that address as the IP address. On some operating
3364 systems (e.g. Solaris), it also passes back the IP address string as the
3365 "host name". However, on others (e.g. Linux), it passes back an empty
3366 string. Exim wasn't checking for this, and was changing the host name to an
3367 empty string, assuming it had been canonicized.
3369 5. Although rare, it is permitted to have more than one PTR record for a given
3370 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3371 all the names associated with an address, because they do in Solaris.
3372 However, it seems that they do not in Linux for data that comes from the
3373 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3374 I found this out when I moved to a new Linux workstation and tried to run
3375 the Exim test suite.
3377 To get round this problem I have changed the code so that it now does its
3378 own call to the DNS to look up PTR records when searching for a host name.
3379 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3380 addresses that are only in /etc/hosts are still found.
3382 This behaviour is, however, controlled by an option called host_lookup_
3383 order, which defaults to "bydns:byaddr". If people want to use the other
3384 order, or indeed, just use one or the other means of lookup, they can
3385 specify it in this variable.
3387 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3388 some operating systems, this comes back from gethostbyaddr() as an empty
3389 string, and this is what Exim used to test for. However, it seems that in
3390 other systems, "." is yielded. Exim now tests for this case too.
3392 7. The values of check_spool_space and check_log_space are now held internally
3393 as a number of kilobytes instead of an absolute number of bytes. If a
3394 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3395 kilobyte. This means that much larger values can be stored.
3397 8. Exim monitor: an attempt to get the action menu when not actually pointing
3398 at a message produces an empty menu entitled "No message selected". This
3399 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3400 no entries in it ("Shell widget menu has zero width and/or height"). So I
3401 have added a single, blank menu entry in this case.
3403 9. Added ${quote_local_part.
3405 10. MIME decoding is now applied to the contents of Subject: header lines when
3408 11. Now that a reference to $sender_host_address automatically causes a reverse
3409 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3410 host lookup before query-style lookups in lists that might use this
3411 variable. This has therefore been abolished, and the "net-" prefix is no
3412 longer necessary for query-style lookups.
3414 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3415 have been a typo for LFLAGS, so it has been changed.
3417 13. The install script calls Exim with "-C /dev/null" in order to find the
3418 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3419 to be output. Howeve, since Exim outputs its version number before the
3420 error, it didn't break the script. It just looked ugly. I fixed this by
3421 always allowing "-C /dev/null" if the caller is root.
3423 14. Ignore overlarge ACL variable number when reading spool file - insurance
3424 against a later release with more variables having written the file.
3426 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3427 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3428 this, as well as the form without the "IPv6" on the front.
3430 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3431 OS/Makefile-Darwin file.
3433 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3434 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3436 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3437 variable. However, some OS use ints for inodes. I've added cast to long int
3438 to get rid of the compiler warning.
3440 19. I had forgotten to lock out "/../" in configuration file names when
3441 ALT_CONFIG_PREFIX was set.
3443 20. Routers used for verification do not need to specify transports. However,
3444 if such a router generated a host list, and callout was configured, Exim
3445 crashed, because it could not find a port number from the (non-existent)
3446 transport. It now assumes port 25 in this circumstance.
3448 21. Added the -t option to exigrep.
3450 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3451 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3452 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3453 them is set, LOOKUP_LSEARCH is forced.
3455 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3457 24. Added sender and host information to the "rejected by local_scan()" log
3458 line; previously there was no indication of these.
3460 25. Added .include_if_exists.
3462 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3463 Linux. It turns out that not all file systems support this. Apparently some
3464 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3465 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3466 not supported on the file descriptor, is now ignored when Exim is trying to
3467 sync a directory. This applies only to Linux.
3469 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3471 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3472 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3475 29. Added support for maildirsize files from supplied patch (modified a bit).
3477 30. The use of :fail: followed by an empty string could lead Exim to respond to
3478 sender verification failures with (e.g.):
3480 550 Verification failed for <xxx>
3481 550 Sender verify failed
3483 where the first response line was missing the '-' that indicates it is not
3484 the final line of the response.
3486 31. The loop for finding the name of the user that called Exim had a hardwired
3487 limit of 10; it now uses the value of finduser_retries, which is used for
3488 all other user lookups.
3490 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3493 33. Exim was neglecting to zero errno before one call of strtol() when
3494 expanding a string and expecting an integer value. On some systems this
3495 resulted in spurious "integer overflow" errors. Also, it was casting the
3496 result into an int without checking.
3498 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3499 did not work. The code looks as if it has *never* worked, though it appears
3500 to have been documented since at least releast 1.62. I have made it work.
3502 35. The "timeout_DNS" error in retry rules, also documented since at least
3503 1.62, also never worked. As it isn't clear exactly what this means, and
3504 clearly it isn't a major issue, I have abolished the feature by treating it
3505 as "timeout", and writing a warning to the main and panic logs.
3507 36. The display of retry rules for -brt wasn't always showing the error code
3510 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3511 timeout_connect_A, timeout_connect_MX.
3513 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3514 to the empty sender.
3516 39. The daemon was not analysing the content of -oX till after it had closed
3517 stderr and disconnected from the controlling terminal. This meant that any
3518 syntax errors were only noted on the panic log, and the return code from
3519 the command was 0. By re-arranging the code a little, I've made the
3520 decoding happen first, so such errors now appear on stderr, and the return
3521 code is 1. However, the actual setting up of the sockets still happens in
3522 the disconnected process, so errors there are still only recorded on the
3525 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3526 connections (as happens on some IP stacks) was logged at start up time as
3527 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3528 it from "IPv6 and IPv4", which means that two separate sockets are being
3531 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3532 says whether AF_INET or AF_INET6 was passed as an argument.
3534 42. Exiwhat output was messed up when time zones were included in log
3537 43. Exiwhat now gives more information about the daemon's listening ports,
3538 and whether -tls-on-connect was used.
3540 44. The "port" option of the smtp transport is now expanded.
3542 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3543 silently ignored. Now an error message is written to the main and panic
3546 46. There's a new ACL modifier called "logwrite" which writes to a log file
3547 as soon as it is encountered.
3549 47. Added $local_user_uid and $local_user_gid at routing time.
3551 48. Exim crashed when trying to verify a sender address that was being
3554 49. Exim was recognizing only a space character after ".include". It now also
3555 recognizes a tab character.
3557 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3558 extracting the relevant information from the specification. The man page no
3559 longer contains scrambled data for the -d option, and I've added a section
3560 at the front about calling Exim under different names.
3562 51. Added "extra_headers" argument to the "mail" command in filter files.
3564 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3567 53. Installed eximstats 1.29.
3569 54. Added transport_filter_timeout as a generic transport option.
3571 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3572 Cc: header lines. This was required by RFC 822, but it not required by RFC
3575 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3576 incoming messages that did not have them. Now it does so only if the
3577 message originates locally, that is, if there is no associated remote host
3578 address. When Resent- header lines are present, this applies to the Resent-
3579 lines rather than the non-Resent- lines.
3581 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3582 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3584 58. Messages for configuration errors now include the name of the main
3585 configuration file - useful now that there may be more than one file in a
3586 list (.included file names were always shown).
3588 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3589 for those rare installations that do not start the daemon as root or run it
3590 setuid root. I've cut out the call to initgroups() if the daemon is not
3593 60. The Exim user and group can now be bound into the binary as text strings
3594 that are looked up at the start of Exim's processing.
3596 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3598 62. Added $mailstore_basename variable.
3600 63. Installed patch to sieve.c from Michael Haardt.
3602 64. When Exim failed to open the panic log after failing to open the main log,
3603 the original message it was trying to log was written to stderr and debug
3604 output, but if they were not available (the usual case in production), it
3605 was lost. Now it is written to syslog before the two lines that record the
3606 failures to open the logs.
3608 65. Users' Exim filters run in subprocesses under the user's uid. It is
3609 possible for a "deliver" command or an alias in a "personal" command to
3610 provoke an address rewrite. If logging of address rewriting is configured,
3611 this fails because the process is not running as root or exim. There may be
3612 a better way of dealing with this, but for the moment (because 4.30 needs
3613 to be released), I have disabled address rewrite logging when running a
3614 filter in a non-root, non-exim process.
3620 1. The buildconfig auxiliary program wasn't quoting the value set for
3621 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3622 not defined. This bug was masked in 4.22 by the effect that was fixed in
3625 2. Some messages that were rejected after a message id was allocated were
3626 shown as "incomplete" by exigrep. It no longer does this for messages that
3627 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3629 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3630 have allow_domain_literals set, the ID did not get logged in the <= line.
3631 Domain literals are now always recognized in Message-ID: header lines.
3633 4. The first argument for a ${extract expansion item is the key name or field
3634 number. Leading and trailing spaces in this item were not being ignored,
3635 causing some misleading effects.
3637 5. When deliver_drop_privilege was set, single queue runner processes started
3638 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3639 same command in the process it spins off) were not dropping privilege.
3641 6. When the daemon running as "exim" started a queue runner, it always
3642 re-executed Exim in the spun-off process. This is a waste of effort when
3643 deliver_drop_privilege is set. The new process now just calls the
3644 queue-runner function directly.
3650 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3653 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3654 ignored. Though the use of -oP was forcing the writing of a pid file, it
3655 was always written to the default place.
3657 3. If the message "no IP address found for host xxxx" is generated during
3658 incoming verification, it is now followed by identification of the incoming
3659 connection (so you can more easily find what provoked it).
3661 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3663 5. Added some features to "harden" Exim a bit more against certain attacks:
3665 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3666 be put in Local/Makefile. This is like the never_users runtime option,
3667 but it cannot be overridden. The default setting is "root".
3669 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3670 prefix string with which any file named in a -C command line option
3673 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3674 is retained for -C and -D only if the caller of Exim is root. Without
3675 it, the exim user may also use -C and -D and retain privilege.
3677 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3678 command line option is disabled.
3680 6. Macro names set by the -D option must start with an upper case letter, just
3681 like macro names defined in the configuration file.
3683 7. Added "dereference=" facility to LDAP.
3685 8. Two instances of the typo "uknown" in the source files are fixed.
3687 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3688 the Configure-Makefile script screwed up while processing it.
3690 10. Incorporated PCRE 4.4.
3692 11. The SMTP synchronization check was not operating right at the start of an
3693 SMTP session. For example, it could not catch a HELO sent before the client
3694 waited for the greeting. There is now a check for outstanding input at the
3695 point when the greeting is written. Because of the duplex, asynchronous
3696 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3697 way, but not yet received, when the check is performed.
3699 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3700 on TCP/IP sockets, because this apparently causes some broken clients to
3703 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3704 unchanged) from the Cygwin maintainer.
3706 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3707 was supported instead of testing for mbx. Effectively a typo.
3709 15. The spa authenticator server code was not checking that the input it
3710 received was valid base64.
3712 16. The debug output line for the "set" modifier in ACLs was not showing the
3713 name of the variable that was being set.
3715 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3717 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3718 lookup on its own. Something else was needed to trigger the lookup. For
3719 example, a match in host_lookup or the need for a host name in a host list.
3720 Now, if $sender_host_name is referenced and the host name has not yet been
3721 looked up, a lookup is performed. If the lookup fails, the variable remains
3722 empty, and $host_lookup_failed is set to "1".
3724 19. Added "eqi" as a case-independent comparison operator.
3726 20. The saslauthd authentication condition could segfault if neither service
3727 nor realm was specified.
3729 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3730 error message that was logged was misleading, and incoming SMTP
3731 connections were dropped. The message is now more accurate, and temporary
3732 errors are given to SMTP connections.
3734 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3735 (except RSET) with a 421 error, until QUIT is received. However, it was
3736 failing to send a response to QUIT.
3738 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3739 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3740 ACL is now run after the helo_try_verify_hosts code.
3742 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3743 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3744 case-independent, so other case variants are also recognized). Apparently
3745 some systems use these upper case variants.
3747 25. If more than two messages were waiting for the same host, and a transport
3748 filter was specified for the transport, Exim sent two messages over the
3749 same TCP/IP connection, and then failed with "socket operation on non-
3750 socket" when it tried to send the third.
3752 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3754 27. The extern definition of crypt16() in expand.c was not being excluded when
3755 the OS had its own crypt16() function.
3757 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3758 as a preferred synonym for return_size_limit, both as an option and as an
3761 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3763 30. Changed the default configuration ACL to relax the local part checking rule
3764 for addresses that are not in any local domains. For these addresses,
3765 slashes and pipe symbols are allowed within local parts, but the sequence
3766 /../ is explicitly forbidden.
3768 31. SPA server authentication was not clearing the challenge buffer before
3771 32. log_message in a "warn" ACL statement was writing to the reject log as
3772 well as to the main log, which contradicts the documentation and doesn't
3773 seem right (because no rejection is happening). So I have stopped it.
3775 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3776 However, I am unable to do any testing of this.
3778 34. Fixed an infelicity in the appendfile transport. When checking directories
3779 for a mailbox, to see if any needed to be created, it was accidentally
3780 using path names with one or more superfluous leading slashes; tracing
3781 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3783 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3784 contents are added to the log line that is written for every discarded
3785 recipient. (Previously a log_message setting was ignored.)
3787 36. The ${quote: operator now quotes the string if it is empty.
3789 37. The install script runs exim in order to find its version number. If for
3790 some reason other than non-existence or emptiness, which it checks, it
3791 could not run './exim', it was installing it with an empty version number,
3792 i.e. as "exim-". This error state is now caught, and the installation is
3795 38. An argument was missing from the function that creates an error message
3796 when Exim fails to connect to the socket for saslauthd authentication.
3797 This could cause Exim to crash, or give a corrupted message.
3799 39. Added isip, isip4, and isip6 to ${if conditions.
3801 40. The ACL variables $acl_xx are now saved with the message, and can be
3802 accessed later in routers, transports, and filters.
3804 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3805 strings in the file are not string-expanded.
3807 42. If a MAIL command specified a SIZE value that was too large to fit into an
3808 int variable, the check against message_size_limit failed. Such values are
3809 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3810 day this will have to be increased, but I don't think I want to be around
3811 when emails are that large.
3818 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3819 iconv() is not standard in FreeBSD.
3821 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3822 IPv6 enabled. The observed symptom was a segmentation fault on return from
3823 the function os_common_find_running_interfaces() in src/os.c.
3825 3. In the check_special_case() function in daemon.c I had used "errno" as an
3826 argument name, which causes warnings on some systems. This was basically a
3827 typo, since it was named "eno" in the comments!
3829 4. The code that waits for the clock to tick (at a resolution of some fraction
3830 of a second) so as to ensure message-id uniqueness was always waiting for
3831 at least one whole tick, when it could have waited for less. [This is
3832 almost certainly not relevant at current processor speeds, where it is
3833 unlikely to ever wait at all. But we try to future-proof.]
3835 5. The function that sleeps for a time interval that includes fractions of a
3836 second contained a race. It did not block SIGALRM between setting the
3837 timer, and suspending (a couple of lines later). If the interval was short
3838 and the sigsuspend() was delayed until after it had expired, the suspension
3839 never ended. On busy systems this could lead to processes getting stuck for
3842 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3843 process, before it forks any delivery processes. The open lookup caching
3844 mechanism meant that the open file or database connection was passed into
3845 the delivery process. The problem was that delivery processes always tidy
3846 up cached lookup data. This could cause a problem for the next delivery
3847 process started by the queue runner, because the external queue runner
3848 process does not know about the closure. So the next delivery process
3849 still has data in the lookup cache. In the case of a file lookup, there was
3850 no problem because closing a file descriptor in a subprocess doesn't affect
3851 the parent. However, if the lookup was caching a connection to a database,
3852 the connection was closed, and the second delivery process was likely to
3853 see errors such as "PGSQL: query failed: server closed the connection
3854 unexpectedly". The problem has been fixed by closing all cached lookups
3855 in a queue runner before running a delivery process.
3857 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3858 seem to have the "const" qualifier which it has on other OS. I've
3861 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3862 *of the same type* (client or server) with the same public name that an
3863 error should be diagnosed.
3865 9. When Exim looked up a host name for an IP address, but failed to find the
3866 original IP address when looking up the host name (a safety check), it
3867 output the message "<ip address> does not match any IP for NULL", which was
3868 confusing, to say the least. The bug was that the host name should have
3869 appeared instead of "NULL".
3871 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3872 user that is built into the binary, and the -C or -D options is used, root
3873 privilege is dropped before the configuration file is read. In addition,
3874 logging is switched to stderr instead of the normal log files. If the
3875 configuration then re-defines the Exim user, the unprivileged environment
3876 is probably not what is expected, so Exim logs a panic warning message (but
3879 However, if deliver_drop_privilege is set, the unprivileged state may well
3880 be exactly what is intended, so the warning has been cut out in that case,
3881 and Exim is allowed to try to write to its normal log files.
3887 1. smtp_return_error_details was not giving details for temporary sender
3888 or receiver verification errors.
3890 2. Diagnose a configuration error if two authenticators have the same public
3893 3. Exim used not to create the message log file for a message until the first
3894 delivery attempt. This could be confusing when incoming messages were held
3895 for policy or load reasons. The message log file is now created at the time
3896 the message is received, and an initial "Received" line is written to it.
3898 4. The automatically generated man page for command line options had a minor
3899 bug that caused no ill effects; however, a more serious problem was that
3900 the procedure for building the man page automatically didn't always
3901 operate. Consequently, release 4.20 contains an out-of-date version. This
3902 shouldn't happen again.
3904 5. When building Exim with embedded Perl support, the script that builds the
3905 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3906 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3908 6. The freeze_tell option was not being used for messages that were frozen on
3909 arrival, either by an ACL or by local_scan().
3911 7. Added the smtp_incomplete_transaction log selector.
3913 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3914 was accepting AUTH without a new EHLO.
3916 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3917 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3920 10. Logging of TCP/IP connections (when configured) now happens in the main
3921 daemon process instead of the child process, so that the TCP/IP connection
3922 count is more accurate (but it can never be perfect).
3924 11. The use of "drop" in a nested ACL was not being handled correctly in the
3925 outer ACL. Now, if condition failure induced by the nested "drop" causes
3926 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3927 or "require"), the connection is dropped.
3929 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3930 that yield "discard" can now be used with an "accept" or a "discard" verb,
3931 but an error is generated for any others (because I can't see a useful way
3932 to define what should happen).
3934 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3935 are now processed in the same way as lines in the Exim configuration file.
3936 In particular, continuation lines are supported.
3938 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3940 15. Added -ti meaning -t -i.
3942 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3943 domains, and warn by logging if others are found.
3945 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3946 returned from GIFCONF when Exim is trying to find the list of interfaces on
3947 a host. The code in os.c has been modified to copy each ifreq to an aligned
3948 structure in all cases.
3950 Also, in some cases, the returned ifreq's were being copied to a 'struct
3951 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3952 means the last couple of bytes of an IPv6 address could be chopped if the
3953 ifreq contained only a normal sockaddr (14 bytes storage).
3955 18. Named domain lists were not supported in the hosts_treat_as_local option.
3956 An entry such as +xxxx was not recognized, and was treated as a literal
3959 19. Ensure that header lines added by a DATA ACL are included in the reject log
3960 if the ACL subsequently rejects the message.
3962 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3963 MD5 (which is deprecated).
3965 21. When testing a filter file using -bf, Exim was writing a message when it
3966 took the sender from a "From " line in the message, but it was not doing so
3967 when it took $return_path from a Return-Path: header line. It now does.
3969 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3970 with a valid header line field name (a series of printing characters
3971 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3973 23. Changed "disc" in the source to "disk" to conform to the documentation and
3974 the book and for uniformity.
3976 24. Ignore Sendmail's -Ooption=value command line item.
3978 25. When execve() failed while trying to run a command in a pipe transport,
3979 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3980 could be confused with a return value of 69 from the command itself. This
3981 has been changed to 127, the value the shell returns if it is asked to run
3982 a non-existent command. The wording for the related log line suggests a
3983 non-existent command as the problem.
3985 26. If received_header_text expands to an empty string, do not add a Received:
3986 header line to the message. (Well, it adds a token one on the spool, but
3987 marks it "old" so that it doesn't get used or transmitted.)
3989 27. Installed eximstats 1.28 (addition of -nt option).
3991 28. There was no check for failure on the call to getsockname() in the daemon
3992 code. This can fail if there is a shortage of resources on the system, with
3993 ENOMEM, for example. A temporary error is now given on failure.
3995 29. Contrary to the C standard, it seems that in some environments, the
3996 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3997 program. Exim now does this explicitly; it affects the formatting of
3998 timestamps using strftime().
4000 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4001 complaints. I've now initialized all the variables, to avoid this.
4003 32. Header lines added by a system filter were not being "seen" during
4004 transport-time rewrites.
4006 33. The info_callback() function passed to OpenSSL is set up with type void
4007 (*)(SSL *, int, int), as described somewhere. However, when calling the
4008 function (actually a macro) that sets it up, the type void(*)() is
4009 expected. I've put in a cast to prevent warnings from picky compilers.
4011 34. If a DNS black list lookup found a CNAME record, but there were no A
4012 records associated with the domain it pointed at, Exim crashed.
4014 35. If a DNS black list lookup returned more than one A record, Exim ignored
4015 all but the first. It now scans all returned addresses if a particular IP
4016 value is being sought. In this situation, the contents of the
4017 $dnslist_value variable are a list of all the addresses, separated by a
4020 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4021 to accept a host name and all its aliases if the forward lookup for any of
4022 them yielded the IP address of the incoming connection. Now it accepts only
4023 those names whose forward lookup yields the correct IP address. Any other
4024 names are discarded. This closes a loophole whereby a rogue DNS
4025 administrator could create reverse DNS records to break through a
4026 wildcarded host restriction in an ACL.
4028 37. If a user filter or a system filter that ran in a subprocess used any of
4029 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4030 the wrong values were passed to the pipe command ($thisaddress had the
4031 value of $0, $0 had the value of $1, etc). This bug was introduced by
4032 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4034 38. Improved the line breaking for long SMTP error messages from ACLs.
4035 Previously, if there was no break point between 40 and 75 characters, Exim
4036 left the rest of the message alone. Two changes have been made: (a) I've
4037 reduced the minimum length to 35 characters; (b) if it can't find a break
4038 point between 35 and 75 characters, it looks ahead and uses the first one
4039 that it finds. This may give the occasional overlong line, but at least the
4040 remaining text gets split now.
4042 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4043 file descriptors might be low, and that setting 1000 would always raise it.
4044 It turns out that in some environments, the limit is already over 1000 and
4045 that lowering it causes trouble. So now Exim takes care not to decrease it.
4047 40. When delivering a message, the value of $return_path is set to $sender_
4048 address at the start of routing (routers may change the value). By an
4049 oversight, this default was not being set up when an address was tested by
4050 -bt or -bv, which affected the outcome if any router or filter referred to
4053 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4054 log only when "message" or "log_message" are set. However, if one of the
4055 conditions was an address verification, or a call to a nested ACL, the
4056 messages generated by the underlying test were being passed through. This
4057 no longer happens. The underlying message is available in $acl_verify_
4058 message for both "message" and "log_message" expansions, so it can be
4059 passed through if needed.
4061 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4062 new expansion $bh_ to give the encoded byte string without charset
4063 translation. Translation happens only if iconv() is available; HAVE_ICONV
4064 indicates this at build time. HEADERS_CHARSET gives the charset to
4065 translate to; headers_charset can change it in the configuration, and
4066 "headers charset" can change it in an individual filter file.
4068 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4069 that creates RFC 2047 encoded "words" labels them as that charset instead
4070 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4071 expansion operator; (ii) when Exim creates a From: line for a local
4072 message; (iii) when a header line is rewritten to include a "phrase" part.
4074 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4075 buggy, causing it to skip the first lines of messages whose message ID
4076 ended in 'D'. This would not have bitten before Exim release 4.14, because
4077 message IDs were unlikely to end in 'D' before then. The effect was to have
4078 incorrect size information for certain domains.
4080 45. #include "config.h" was missing at the start of the crypt16.c module. This
4081 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4084 46. If there was a timeout during a "random" callout check, Exim treated it as
4085 a failure of the random address, and carried on sending RSET and the real
4086 address. If the delay was just some slowness somewhere, the response to the
4087 original RCPT would be taken as a response to RSET and so on, causing
4088 mayhem of various kinds.
4090 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4091 when I implemented it. It didn't allow for the fact that some option values
4092 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4093 the right test for positive values.
4095 48. Domain names in DNS records are case-independent. Exim always looks them up
4096 in lower case. Some resolvers return domain names in exactly the case they
4097 appear in the zone file, that is, they may contain uppercase letters. Not
4098 all resolvers do this - some return always lower case. Exim was treating a
4099 change of case by a resolver as a change of domain, similar to a widening
4100 of a domain abbreviation. This triggered its re-routing code and so it was
4101 trying to route what was effectively the same domain again. This normally
4102 caused routing to fail (because the router wouldn't handle the domain
4103 twice). Now Exim checks for this case specially, and just changes the
4104 casing of the domain that it ultimately uses when it transmits the message
4107 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4110 50. If a filter generated a file delivery with a non-absolute name (possible if
4111 no home directory exists for the router), the forbid_file option was not
4114 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4115 the existing equality matching.
4117 52. Exim was using ints instead of ino_t variables in some places where it was
4118 dealing with inode numbers.
4120 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4121 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4122 called TMPDIR, and if it finds it is different, it changes its value.
4124 54. The smtp_printf() function is now made available to local_scan() so
4125 additional output lines can be written before returning. There is also an
4126 smtp_fflush() function to enable the detection of a dropped connection.
4127 The variables smtp_input and smtp_batched_input are exported to
4130 55. Changed the default runtime configuration: the message "Unknown user"
4131 has been removed from the ACL, and instead placed on the localuser router,
4132 using the cannot_route_message feature. This means that any verification
4133 failures that generate their own messages won't get overridden. Similarly,
4134 the "Unrouteable address" message that was in the ACL for unverifiable
4135 relay addresses has also been removed.
4137 56. Added hosts_avoid_esmtp to the smtp transport.
4139 57. The exicyclog script was not checking for the esoteric option
4140 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4141 will work only if exicyclog is run under the appropriate euid.
4143 58. Following a discussion on the list, the rules by which Exim recognises line
4144 endings on incoming messages have been changed. The -dropcr and drop_cr
4145 options are now no-ops, retained only for backwards compatibility. The
4146 following line terminators are recognized: LF CRLF CR. However, special
4147 processing applies to CR:
4149 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4150 nor a local message in the state where . is a terminator.
4152 (ii) If a bare CR is encountered in a header line, an extra space is added
4153 after the line terminator so as not to end the header. The reasoning
4154 behind this is that bare CRs in header lines are most likely either
4155 to be mistakes, or people trying to play silly games.
4157 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4158 was being incorrectly given as 18 bytes larger than it should have been.
4159 This is a VOB (very old bug).
4161 60. This may never have affected anything current, but just in case it has:
4162 When the local host is found other than at the start of a list of hosts,
4163 the local host, those with the same MX, and any that follow, are discarded.
4164 When the list in question was part of a longer list of hosts, the following
4165 hosts (not currently being processed) were also being discarded. This no
4166 longer happens. I'm not sure if this situation could ever has previously
4169 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4172 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4173 header. This is something that is recommended in a new Internet Draft, and
4174 is something that is documented as being done by Sendmail. There are two
4175 possible values. For messages generated by the autoreply transport, Exim
4178 Auto-Submitted: auto-replied
4180 whereas for all other generated messages (e.g. bounces) it adds
4182 Auto-Submitted: auto-generated
4184 63. The "personal" condition in filters now includes a test for the
4185 Auto-Submitted: header. If it contains the string "auto-" the message it
4186 not considered personal.
4188 64. Added rcpt_include_affixes as a generic transport option.
4190 65. Added queue_only_override (default true).
4192 66. Added the syslog_duplication option.
4194 67. If what should have been the first header line of a message consisted of
4195 a space followed by a colon, Exim was mis-interpreting it as a header line.
4196 It isn't of course - it is syntactically invalid and should therefore be
4197 treated as the start of the message body. The misbehaviour could have
4198 caused a number of strange effects, including loss of data in subsequent
4199 header lines, and spool format errors.
4201 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4202 client host had authenticated. This control can now be exercised by an ACL
4203 for more flexibility.
4205 69. By default, callouts do not happen when testing with -bh. There is now a
4206 variant, -bhc, which does actually run the callout code, including
4207 consulting and updating the callout cache.
4209 70. Added support for saslauthd authentication, courtesy of Alexander
4212 71. If statvfs() failed on the spool or log directories while checking their
4213 size for availability, Exim confusingly gave the error "space shortage".
4214 Furthermore, in debugging mode it crashed with a floating point exception.
4215 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4216 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4217 serious problem, Exim now writes to the main and panic logs when this
4218 happens, with details of the failure. It then refuses to accept the
4219 incoming message, giving the message "spool directory problem" or "log
4220 directory problem" with a 421 code for SMTP messages.
4222 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4223 0, 1, and 2 exist, because some OS complain for execs without them (see
4224 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4225 descriptors. However, the code omitted to check that the open succeeded,
4226 causing mysterious errors if for some reason the permissions on /dev/null
4227 got screwed. Now Exim writes a message to the main and panic logs, and
4228 bombs out if it can't open /dev/null.
4230 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4231 interact so that it is all more flexible. It is supposed to remain
4232 backwards compatible. Also added extra_local_interfaces.
4234 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4235 to bomb out with an assertion failure - to the client this appears as a
4236 connection drop. This problem occurs in the part of the code that was taken
4237 from the Samba project. Fortunately, the assertion is in a very simple
4238 function, so I have fixed this by reproducing the function inline in the
4239 one place where it is called, and arranging for authentication to fail
4240 instead of killing the process with assert().
4242 75. The SPA client code was not working when the server requested OEM rather
4243 than Unicode encoding.
4245 76. Added code to make require_files with a specific uid setting more usable in
4246 the case where statting the file as root fails - usually a non-root-mounted
4247 NFS file system. When this happens and the failure is EACCES, Exim now
4248 forks a subprocess and does the per-uid checking as the relevant uid.
4250 77. Added process_log_path.
4252 78. If log_file_path was not explicitly set, a setting of check_log_space or
4253 check_log_inodes was ignored.
4255 79. If a space check for the spool or log partitions fails, the incident is now
4256 logged. Of course, in the latter case the data may get lost...
4258 80. Added the %p formatting code to string_format() so that it can be used to
4259 print addresses in debug_print(). Adjusted all the address printing in the
4260 debugging in store.c to use %p rather than %d.
4262 81. There was a concern that a line of code in smtp_in.c could overflow a
4263 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4264 initially expressed, the concern was not well-founded, because trailing
4265 spaces are removed early. However, if the trailing spaces were followed by
4266 a NULL, they did not get removed, so the overflow was possible. Two fixes
4269 (a) I re-wrote the offending code in a cleaner fashion.
4270 (b) If an incoming SMTP command contains a NULL character, it is rejected
4273 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4274 runs initgroups(), so that if the Exim user is in any additional groups,
4275 they will be used during message reception.
4281 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob