Kjetil Homme's patch for extended macro features (redefinition,
[exim.git] / doc / doc-txt / ChangeLog
cf00dad6 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.108 2005/04/05 13:58:34 ph10 Exp $
3Change log file for Exim from version 4.21
bbe902f0 6
7Exim version 4.51
10TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
11 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2f079f46 13TK/02 Fix ACL "control" statement not being available in MIME ACL.
15TK/03 Fix ACL "regex" condition not being available in MIME ACL.
17PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
18 to test Sieve filters that use "vacation".
20PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
21 that changes the way the GnuTLS parameters are stored in the cache file.
22 The new format can be generated externally. For backward compatibility,
23 if the data in the cache doesn't make sense, Exim assumes it has read an
24 old-format file, and it generates new data and writes a new file. This
25 means that you can't go back to an older release without removing the
26 file.
28PH/03 A redirect router that has both "unseen" and "one_time" set does not
29 work if there are any delivery delays because "one_time" forces the
30 parent to be marked "delivered", so its unseen clone is never tried
31 again. For this reason, Exim now forbids the simultaneous setting of
32 these two options.
34PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
35 redirected to themselves ("homonym" addresses). Read the long ChangeLog
36 entry if you want to know the details. The fix, however, neglected to
37 consider the case when local delivery batching is involved. The test for
38 "previously delivered" was not happening when checking to see if an
39 address could be batched with a previous (undelivered) one; under
40 certain circumstances this could lead to multiple deliveries to the same
41 address. A one-line patch to add the appropriate test fixes the bug.
43PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
44 in its include files, and this causes problems building Exim.
46PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
47 header_syntax) but Exim was just ignoring anything given after a slash.
48 In particular, this caused confusion with an attempt to use "verify =
49 reverse_host_lookup/defer_ok". An error is now given when options are
50 supplied for verify items that do not have them. (Maybe reverse_host_
51 lookup should have a defer_ok option, but that's a different point.)
53PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
54 defined by RFC 821) to 2048, because there were problems with some AUTH
55 commands, and RFC 1869 says the size should be increased for extended
56 SMTP commands that take arguments.
58PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
59 Finch).
61PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
62 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 63
64PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
65 an address' error message when a string expansion fails (syntax or
66 whatever). Otherwise not only does the password appear in the log, it may
67 also be put in a bounce message.
69PH/11 Installed exipick version 20050225.0 from John Jetmore.
71PH/12 If the last host in a fallback_hosts list was multihomed, only the first
72 of its addresses was ever tried. (Bugzilla bug #2.)
74PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
75 the result incorrectly in the debug output. (It correctly added a newline
76 to what was transported.)
78TF/01 Added $received_time.
80PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
81 commented out examples of how to interface to a virus scanner and to
82 SpamAssassin. Also added commented examples of av_scanner and
83 spamd_address settings.
85PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
86 and controls are allowed in which ACLs. There were a couple of minor
87 errors. Some of the entries in the conditions table (which is a table of
88 where they are NOT allowed) were getting very unwieldy; rewrote them as a
89 negation of where the condition IS allowed.
91PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
93PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
94 header file does not have a version number, so I've had to invent a new
95 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
96 API. The code is untested by me (my Linux distribution still has 0.3.2 of
97 radiusclient), but it was contributed by a Radius user.
99PH/18 Installed Lars Mainka's patch for the support of CRL collections in
100 files or directories, for OpenSSL.
102PH/19 When an Exim process that is running as root has to create an Exim log
103 file, it does so in a subprocess that runs as exim:exim so as to get the
104 ownership right at creation (otherwise, other Exim processes might see
105 the file with the wrong ownership). There was no test for failure of this
106 fork() call, which would lead to the process getting stuck as it waited
107 for a non-existent subprocess. Forks do occasionally fail when resources
108 run out. I reviewed all the other calls to fork(); they all seem to check
109 for failure.
111PH/20 When checking for unexpected SMTP input at connect time (before writing
112 the banner), Exim was not dealing correctly with a non-positive return
113 from the read() function. If the client had disconnected by this time,
114 the result was a log entry for a synchronization error with an empty
115 string after "input=" when read() returned zero. If read() returned -1
116 (an event I could not check), uninitialized data bytes were printed.
117 There were reports of junk text (parts of files, etc) appearing after
118 "input=".
120PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
122PH/22 Added support for macro redefinition, and (re)definition in between
123 driver and ACL definitions.
7982096b 125
126A note about Exim versions 4.44 and 4.50
129Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
130changes of various kinds. As a consequence, a big documentation update was
131needed. This delayed the release for rather longer than seemed good, especially
132in the light of a couple of (minor) security issues. Therefore, the changes
133that fixed bugs were backported into 4.43, to create a 4.44 maintenance
134release. So 4.44 and 4.50 are in effect two different branches that both start
135from 4.43.
137I have left the 4.50 change log unchanged; it contains all the changes since
1384.43. The change log for 4.44 is below; many of its items are identical to
139those for 4.50. This seems to be the most sensible way to preserve the
140historical information.
f7b63901 143Exim version 4.50
146 1. Minor wording change to the doc/README.SIEVE file.
139059f6 148 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 149 computation of the current number of files was incorrect.
495ae4b0 150
151 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
152 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
153 place.
155 4. Give more explanation in the error message when the command for a transport
156 filter fails to execute.
158 5. There are several places where Exim runs a non-Exim command in a
159 subprocess. The SIGUSR1 signal should be disabled for these processes. This
160 was being done only for the command run by the queryprogram router. It is
161 now done for all such subprocesses. The other cases are: ${run, transport
162 filters, and the commands run by the lmtp and pipe transports.
164 6. Added CONFIGURE_GROUP build-time option.
166 7. Some older OS have a limit of 256 on the maximum number of file
167 descriptors. Exim was using setrlimit() to set 1000 as a large value
168 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
169 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 170
171 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
172 was an oversight, and furthermore, ever since the addition of extra
173 controls (e.g. 4.43/32), the checks on when to allow different forms of
174 "control" were broken. There should now be diagnostics for all cases when a
175 control that does not make sense is encountered.
177 9. Added the /retain_sender option to "control=submission".
17910. $recipients is now available in the predata ACL (oversight).
18111. Tidy the search cache before the fork to do a delivery from a message
182 received from the command line. Otherwise the child will trigger a lookup
183 failure and thereby defer the delivery if it tries to use (for example) a
184 cached ldap connection that the parent has called unbind on.
18612. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
187 of $address_data from the recipient verification was clobbered by the
188 sender verification.
19013. The value of address_data from a sender verification is now available in
191 $sender_address_data in subsequent conditions in the ACL statement.
19314. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
19515. Added a new option "connect=<time>" to callout options, to set a different
196 connection timeout.
19816. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
199 was its contents. (It was OK if the option was not defined at all.)
20117. A "Completed" log line is now written for messages that are removed from
202 the spool by the -Mrm option.
20418. New variables $sender_verify_failure and $recipient_verify_failure contain
205 information about exactly what failed.
20719. Added -dd to debug only the daemon process.
20920. Incorporated Michael Haardt's patch to ldap.c for improving the way it
210 handles timeouts, both on the server side and network timeouts. Renamed the
211 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
21321. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
214 It is now set to "smtps".
21622. $host_address is now set to the target address during the checking of
217 ignore_target_hosts.
21923. When checking ignore_target_hosts for an ipliteral router, no host name was
220 being passed; this would have caused $sender_host_name to have been used if
221 matching the list had actually called for a host name (not very likely,
222 since this list is usually IP addresses). A host name is now passed as
223 "[x.x.x.x]".
22524. Changed the calls that set up the SIGCHLD handler in the daemon to use the
226 code that specifies a non-restarting handler (typically sigaction() in
227 modern systems) in an attempt to fix a rare and obscure crash bug.
22925. Narrowed the window for a race in the daemon that could cause it to ignore
230 SIGCHLD signals. This is not a major problem, because they are used only to
231 wake it up if nothing else does.
23326. A malformed maildirsize file could cause Exim to calculate negative values
234 for the mailbox size or file count. Odd effects could occur as a result.
235 The maildirsize information is now recalculated if the size or filecount
236 end up negative.
23827. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
239 support for a long time. Removed HAVE_SYS_VFS_H.
24128. Installed the latest version of exipick from John Jetmore.
24329. In an address list, if the pattern was not a regular expression, an empty
244 subject address (from a bounce message) matched only if the pattern was an
245 empty string. Non-empty patterns were not even tested. This was the wrong
246 because it is perfectly reasonable to use an empty address as part of a
247 database query. An empty address is now tested by patterns that are
248 lookups. However, all the other forms of pattern expect the subject to
249 contain a local part and a domain, and therefore, for them, an empty
250 address still always fails if the pattern is not itself empty.
25230. Exim went into a mad DNS loop when attempting to do a callout where the
253 host was specified on an smtp transport, and looking it up yielded more
254 than one IP address.
25631. Re-factored the code for checking spool and log partition space into a
257 function that finds that data and another that does the check. The former
258 is then used to implement four new variables: $spool_space, $log_space,
259 $spool_inodes, and $log_inodes.
26132. The RFC2047 encoding function was originally intended for short strings
262 such as real names; it was not keeping to the 75-character limit for
263 encoded words that the RFC imposes. It now respects the limit, and
264 generates multiple encoded words if necessary. To be on the safe side, I
265 have increased the buffer size for the ${rfc2047: expansion operator from
266 1024 to 2048 bytes.
26833. It is now permitted to omit both strings after an "if" condition; if the
269 condition is true, the result is "true". As before, when the second string
270 is omitted, a false condition yields an empty string. This makes it less
271 cumbersome to write custom ACL and router conditions.
27334. Failure to deliver a bounce message always caused it to be frozen, even if
274 there was an errors_to setting on the router. The errors_to setting is now
275 respected.
27735. If an IPv6 address is given for -bh or -bhc, it is now converted to the
278 canonical form (fully expanded) before being placed in
279 $sender_host_address.
28136. The table in the code that translates DNS record types into text (T_A to
282 "A" for instance) was missing entries for NS and CNAME. It is just possible
283 that this could have caused confusion if both these types were looked up
284 for the same domain, because the text type is used as part of Exim's
285 per-process caching. But the chance of anyone hitting this buglet seems
286 very small.
28837. The dnsdb lookup has been extended in a number of ways.
290 (1) There is a new type, "zns", which walks up the domain tree until it
291 finds some nameserver records. It should be used with care.
293 (2) There is a new type, "mxh", which is like "mx" except that it returns
294 just the host names, not the priorities.
296 (3) It is now possible to give a list of domains (or IP addresses) to be
297 looked up. The behaviour when one of the lookups defers can be
298 controlled by a keyword.
7bb56e1f 299
ea3bc19b 300 (4) It is now possible to specify the separator character for use when
7bb56e1f 301 multiple records are returned.
33397d19 302
30338. The dnslists ACL condition has been extended: it is now possible to supply
304 a list of IP addresses and/or domains to be looked up in a particular DNS
305 domain.
30739. Added log_selector=+queue_time_overall.
30940. When running the queue in the test harness, wait just a tad after forking a
310 delivery process, to get repeatability of debugging output.
31241. Include certificate and key file names in error message when GnuTLS fails
313 to set them up, because the GnuTLS error message doesn't include the name
314 of the failing file when there is a problem reading it.
31642. Allow both -bf and -bF in the same test run.
31843. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
32044. The "Exiscan patch" is now merged into the mainline Exim source.
32245. Sometimes the final signoff response after QUIT could fail to get
323 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
324 < 0 before doing a fflush(). This bug looks as though it goes back to the
325 introduction of TLS in release 3.20, but "sometimes" must have been rare
326 because the tests only now provoked it.
32846. Reset the locale to "C" after calling embedded Perl, in case it was changed
329 (this can affect the format of dates).
33147. exim_tidydb, when checking for the continued existence of a message for
332 which it has found a message-specific retry record, was not finding
333 messages that were in split spool directories. Consequently, it was
334 deleting retry records that should have stayed in existence.
33648. Steve fixed some bugs in eximstats.
33849. The SPA authentication driver was not abandoning authentication and moving
339 on to the next authenticator when an expansion was forced to fail,
340 contradicting the general specification for all authenticators. Instead it
341 was generating a temporary error. It now behaves as specified.
34350. The default ordering of permitted cipher suites for GnuTLS was pessimal
344 (the order specifies the preference for clients). The order is now AES256,
345 AES128, 3DES, ARCFOUR128.
34751. Small patch to Sieve code - explicitly set From: when generating an
348 autoreply.
35052. Exim crashed if a remote delivery caused a very long error message to be
351 recorded - for instance if somebody sent an entire SpamAssassin report back
352 as a large number of 550 error lines. This bug was coincidentally fixed by
353 increasing the size of one of Exim's internal buffers (big_buffer) that
354 happened as part of the Exiscan merge. However, to be on the safe side, I
355 have made the code more robust (and fixed the comments that describe what
356 is going on).
35853. Now that there can be additional text after "Completed" in log lines (if
359 the queue_time_overall log selector is set), a one-byte patch to exigrep
360 was needed to allow it to recognize "Completed" as not the last thing in
361 the line.
36354. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
364 patch that reportedly fixes this has been added. I am not expert enough to
365 create a test for it. This is what the patch creator wrote:
367 "I found a little strange behaviour of ldap code when working with
368 Windows 2003 AD Domain, where users was placed in more than one
369 Organization Units. When I tried to give exim partial DN, the exit code
370 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
371 But simultaneously result of request was absolutely normal ldap result,
372 so I produce this patch..."
374 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
375 REFERENCE, so I have modified the code to exclude the patch when that macro
376 is not defined.
37855. Some experimental protocols are using DNS PTR records for new purposes. The
379 keys for these records are domain names, not reversed IP addresses. The
380 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
381 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 382 CAN-2005-0021
7102e136 383
38456. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
38657. Double the size of the debug message buffer (to 2048) so that more of very
387 long debug lines gets shown.
38958. The exicyclog utility now does better if the number of log files to keep
390 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
39259. Two changes related to the smtp_active_hostname option:
394 (1) $smtp_active_hostname is now available as a variable.
395 (2) The default for smtp_banner uses $smtp_active_hostname instead
396 of $primary_hostname.
39860. The host_aton() function is supposed to be passed a string that is known
399 to be a valid IP address. However, in the case of IPv6 addresses, it was
400 not checking this. This is a hostage to fortune. Exim now panics and dies
401 if the condition is not met. A case was found where this could be provoked
402 from a dnsdb PTR lookup with an IPv6 address that had more than 8
403 components; fortuitously, this particular loophole had already been fixed
404 by change 4.50/55 above.
406 If there are any other similar loopholes, the new check in host_aton()
407 itself should stop them being exploited. The report I received stated that
408 data on the command line could provoke the exploit when Exim was running as
409 exim, but did not say which command line option was involved. All I could
410 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
411 running as the user.
ea3a6f44 412 CAN-2005-0021
41461. There was a buffer overflow vulnerability in the SPA authentication code
415 (which came originally from the Samba project). I have added a test to the
416 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 417 CAN-2005-0022
b975ba52 418
41962. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
420 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
42263. The daemon start-up calls getloadavg() while still root for those OS that
423 need the first call to be done as root, but it missed one case: when
424 deliver_queue_load_max is set with deliver_drop_privilege. This is
425 necessary for the benefit of the queue runner, because there is no re-exec
426 when deliver_drop_privilege is set.
42864. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
429 This has been fixed.
43165. Caching of lookup data for "hosts =" ACL conditions, when a named host list
432 was in use, was not putting the data itself into the right store pool;
433 consequently, it could be overwritten for a subsequent message in the same
434 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
435 the caching.)
43766. Added hosts_max_try_hardlimit to the smtp transport, default 50.
43967. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
440 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
441 the function were treating the return as a boolean value, which happened to
442 work because 0=false and not-0=true, but is not correct code.
44468. The host_aton() function was not handling scoped IPv6 addresses (those
445 with, for example, "%eth0" on the end) correctly.
44769. Fixed some compiler warnings in acl.c for the bitmaps specified with
448 negated items (that is, ~something) in unsigned ints. Some compilers
449 apparently mutter when there is no cast.
45170. If an address verification called from an ACL failed, and did not produce a
452 user-specific message (i.e. there was only a "system" message), nothing was
453 put in $acl_verify_message. In this situation, it now puts the system
454 message there.
45671. Change 4.23/11 added synchronization checking at the start of an SMTP
457 session; change 4.31/43 added the unwanted input to the log line - except
458 that it did not do this in the start of session case. It now does.
46072. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
461 This is wrong and can cause the other end to generate a synchronization
462 error if it is another Exim or anything else that does the synchronization
463 check. A QUIT command is no longer sent after a timeout.
46573. $host_lookup_deferred has been added, to make it easier to detect DEFERs
466 during host lookups.
46874. The defer_ok option of callout verification was not working if it was used
469 when verifying addresses in header lines, that is, for this case:
471 verify = header_sender/callout=defer_ok
47375. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
474 those file descriptors could be used for SMTP connections. If anything
475 wrote to stderr (the example that came up was "warn" in embedded Perl), it
476 could be sent to the SMTP client, causing chaos. The daemon now opens
477 stdin, stdout, and stderr to /dev/null when it puts itself into the
478 background.
48076. Arrange for output from Perl's "warn" command to be written to Exim's main
481 log by default. The user can override this with suitable Perl magic.
48377. The use of log_message on a "discard" ACL verb, which is supposed to add to
484 the log message when discard triggers, was not working for the DATA ACL or
485 for the non-SMTP ACL.
48778. Error message wording change in sieve.c.
48979. If smtp_accept_max_per_host was set, the number of connections could be
490 restricted to fewer than expected, because the daemon was trying to set up
491 a new connection before checking whether the processes handling previous
492 connections had finished. The check for completed processes is now done
493 earlier. On busy systems, this bug wouldn't be noticed because something
494 else would have woken the daemon, and it would have reaped the completed
495 process earlier.
49780. If a message was submitted locally by a user whose login name contained one
498 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
499 It caused a spool format error. I have fixed the spool reading code. A
500 related problem was that the "from" clause in the Received: line became
501 illegal because of the space(s). It is now covered by ${quote_local_part.
50381. Included the latest eximstats from Steve (adds average sizes to HTML Top
504 tables).
50682. Updated OS/Makefile-AIX as per message from Mike Meredith.
50883. Patch from Sieve maintainer to fix unterminated string problem in
509 "vacation" handling.
51184. Some minor changes to the Linux configuration files to help with other
512 OS variants using glibc.
51485. One more patch for Sieve to update vacation handling to latest spec.
495ae4b0 516
518See the note above about the 4.44 and 4.50 releases.
522Exim version 4.44
525 1. Change 4.43/35 introduced a bug that caused file counts to be
526 incorrectly computed when quota_filecount was set in an appendfile
527 transport
529 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
530 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
531 place.
533 3. Give more explanation in the error message when the command for a transport
534 filter fails to execute.
536 4. There are several places where Exim runs a non-Exim command in a
537 subprocess. The SIGUSR1 signal should be disabled for these processes. This
538 was being done only for the command run by the queryprogram router. It is
539 now done for all such subprocesses. The other cases are: ${run, transport
540 filters, and the commands run by the lmtp and pipe transports.
542 5. Some older OS have a limit of 256 on the maximum number of file
543 descriptors. Exim was using setrlimit() to set 1000 as a large value
544 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
545 systems. I've change it so that if it can't get 1000, it tries for 256.
547 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
548 was an oversight, and furthermore, ever since the addition of extra
549 controls (e.g. 4.43/32), the checks on when to allow different forms of
550 "control" were broken. There should now be diagnostics for all cases when a
551 control that does not make sense is encountered.
553 7. $recipients is now available in the predata ACL (oversight).
555 8. Tidy the search cache before the fork to do a delivery from a message
556 received from the command line. Otherwise the child will trigger a lookup
557 failure and thereby defer the delivery if it tries to use (for example) a
558 cached ldap connection that the parent has called unbind on.
560 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
561 of $address_data from the recipient verification was clobbered by the
562 sender verification.
56410. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
565 was its contents. (It was OK if the option was not defined at all.)
56711. A "Completed" log line is now written for messages that are removed from
568 the spool by the -Mrm option.
57012. $host_address is now set to the target address during the checking of
571 ignore_target_hosts.
57313. When checking ignore_target_hosts for an ipliteral router, no host name was
574 being passed; this would have caused $sender_host_name to have been used if
575 matching the list had actually called for a host name (not very likely,
576 since this list is usually IP addresses). A host name is now passed as
577 "[x.x.x.x]".
57914. Changed the calls that set up the SIGCHLD handler in the daemon to use the
580 code that specifies a non-restarting handler (typically sigaction() in
581 modern systems) in an attempt to fix a rare and obscure crash bug.
58315. Narrowed the window for a race in the daemon that could cause it to ignore
584 SIGCHLD signals. This is not a major problem, because they are used only to
585 wake it up if nothing else does.
58716. A malformed maildirsize file could cause Exim to calculate negative values
588 for the mailbox size or file count. Odd effects could occur as a result.
589 The maildirsize information is now recalculated if the size or filecount
590 end up negative.
59217. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
593 support for a long time. Removed HAVE_SYS_VFS_H.
ea3a6f44 59518. Updated exipick to current release from John Jetmore.
59719. Allow an empty sender to be matched against a lookup in an address list.
598 Previously the only cases considered were a regular expression, or an
599 empty pattern.
60120. Exim went into a mad DNS lookup loop when doing a callout where the
602 host was specified on the transport, if the DNS lookup yielded more than
603 one IP address.
60521. The RFC2047 encoding function was originally intended for short strings
606 such as real names; it was not keeping to the 75-character limit for
607 encoded words that the RFC imposes. It now respects the limit, and
608 generates multiple encoded words if necessary. To be on the safe side, I
609 have increased the buffer size for the ${rfc2047: expansion operator from
610 1024 to 2048 bytes.
bbe902f0 611
61222. Failure to deliver a bounce message always caused it to be frozen, even if
613 there was an errors_to setting on the router. The errors_to setting is now
614 respected.
61623. If an IPv6 address is given for -bh or -bhc, it is now converted to the
617 canonical form (fully expanded) before being placed in
618 $sender_host_address.
62024. Updated eximstats to version 1.33
62225. Include certificate and key file names in error message when GnuTLS fails
623 to set them up, because the GnuTLS error message doesn't include the name
624 of the failing file when there is a problem reading it.
62626. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 627 As per change 25.
bbe902f0 628
62927. Reset the locale to "C" after calling embedded Perl, in case it was changed
630 (this can affect the format of dates).
bbe902f0 631
63228. exim_tidydb, when checking for the continued existence of a message for
633 which it has found a message-specific retry record, was not finding
634 messages that were in split spool directories. Consequently, it was
635 deleting retry records that should have stayed in existence.
63729. eximstats updated to version 1.35
638 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
639 1.35 - bugfix such that pie charts by volume are generated correctly
64130. The SPA authentication driver was not abandoning authentication and moving
642 on to the next authenticator when an expansion was forced to fail,
643 contradicting the general specification for all authenticators. Instead it
644 was generating a temporary error. It now behaves as specified.
bbe902f0 645
64631. The default ordering of permitted cipher suites for GnuTLS was pessimal
647 (the order specifies the preference for clients). The order is now AES256,
648 AES128, 3DES, ARCFOUR128.
bbe902f0 649
65031. Small patch to Sieve code - explicitly set From: when generating an
651 autoreply.
bbe902f0 652
65332. Exim crashed if a remote delivery caused a very long error message to be
654 recorded - for instance if somebody sent an entire SpamAssassin report back
655 as a large number of 550 error lines. This bug was coincidentally fixed by
656 increasing the size of one of Exim's internal buffers (big_buffer) that
657 happened as part of the Exiscan merge. However, to be on the safe side, I
658 have made the code more robust (and fixed the comments that describe what
659 is going on).
bbe902f0 660
66133. Some experimental protocols are using DNS PTR records for new purposes. The
662 keys for these records are domain names, not reversed IP addresses. The
663 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
664 leaves it alone. Component reversal etc. now happens only for IP addresses.
665 CAN-2005-0021
66734. The host_aton() function is supposed to be passed a string that is known
668 to be a valid IP address. However, in the case of IPv6 addresses, it was
669 not checking this. This is a hostage to fortune. Exim now panics and dies
670 if the condition is not met. A case was found where this could be provoked
671 from a dnsdb PTR lookup with an IPv6 address that had more than 8
672 components; fortuitously, this particular loophole had already been fixed
673 by change 4.50/55 or 4.44/33 above.
675 If there are any other similar loopholes, the new check in host_aton()
676 itself should stop them being exploited. The report I received stated that
677 data on the command line could provoke the exploit when Exim was running as
678 exim, but did not say which command line option was involved. All I could
679 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
680 running as the user.
681 CAN-2005-0021
68335. There was a buffer overflow vulnerability in the SPA authentication code
684 (which came originally from the Samba project). I have added a test to the
685 spa_base64_to_bits() function which I hope fixes it.
686 CAN-2005-0022
68836. The daemon start-up calls getloadavg() while still root for those OS that
689 need the first call to be done as root, but it missed one case: when
690 deliver_queue_load_max is set with deliver_drop_privilege. This is
691 necessary for the benefit of the queue runner, because there is no re-exec
692 when deliver_drop_privilege is set.
bbe902f0 693
69437. Caching of lookup data for "hosts =" ACL conditions, when a named host list
695 was in use, was not putting the data itself into the right store pool;
696 consequently, it could be overwritten for a subsequent message in the same
697 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
698 the caching.)
bbe902f0 699
70038. Sometimes the final signoff response after QUIT could fail to get
701 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
702 < 0 before doing a fflush(). This bug looks as though it goes back to the
703 introduction of TLS in release 3.20, but "sometimes" must have been rare
704 because the tests only now provoked it.
707Exim version 4.43
710 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
711 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
712 Now it returns an integer. A number of calls were still expecting a T/F
713 return. Fortuitously, in all cases, the tests worked in OK situations,
714 which is the norm. However, things would have gone wrong on any write
715 failures on the smtp file descriptor. This function is used when sending
716 messages over SMTP and also when doing verify callouts.
718 2. When Exim is called to do synchronous delivery of a locally submitted
719 message (the -odf or -odi options), it no longer closes stderr before doing
720 the delivery.
722 3. Implemented the mua_wrapper option.
724 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
726 5. Implemented the functions header_remove(), header_testname(),
727 header_add_at_position(), and receive_remove_recipient(), and exported them
728 to local_scan().
730 6. If an ACL "warn" statement specified the addition of headers, Exim already
731 inserted X-ACL-Warn: at the start if there was no header name. However, it
732 was not making this test for the second and subsequent header lines if
733 there were newlines in the string. This meant that an invalid header could
734 be inserted if Exim was badly configured.
736 7. Allow an ACL "warn" statement to add header lines at the start or after all
737 the Received: headers, as well as at the end.
739 8. Added the rcpt_4xx retry error code.
741 9. Added postmaster_mailfrom=xxx to callout verification option.
74310. Added mailfrom=xxxx to the callout verification option, for verify=
744 header_sender only.
74611. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
747 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
74912. Inserted some casts to stop certain compilers warning when using pointer
750 differences as field lengths or precisions in printf-type calls (mostly
751 affecting debugging statements).
75313. Added optional readline() support for -be (dynamically loaded).
75514. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
756 same clock tick as a message's arrival, so that its received time was the
757 same as the "first fail" time on the retry record, and that message
758 remained on the queue past the ultimate address timeout, every queue runner
759 would try a delivery (because it was past the ultimate address timeout) but
760 after another failure, the ultimate address timeout, which should have then
761 bounced the address, did not kick in. This was a "< instead of <=" error;
762 in most cases the first failure would have been in the next clock tick
763 after the received time, and all would be well.
76515. The special items beginning with @ in domain lists (e.g. @mx_any) were not
766 being recognized when the domain list was tested by the match_domain
767 condition in an expansion string.
76916. Added the ${str2b64: operator.
77117. Exim was always calling setrlimit() to set a large limit for the number of
772 processes, without checking whether the existing limit was already
773 adequate. (It did check for the limit on file descriptors.) Furthermore,
774 errors from getrlimit() and setrlimit() were being ignored. Now they are
775 logged to the main and panic logs, but Exim does carry on, to try to do its
776 job under whatever limits there are.
77818. Imported PCRE 5.0.
78019. Trivial typo in log message " temporarily refused connection" (the leading
781 space).
78320. If the log selector return_path_on_delivery was set and an address was
784 redirected to /dev/null, the delivery process crashed because it assumed
785 that a return path would always be set for a "successful" delivery. In this
786 case, the whole delivery is bypassed as an optimization, and therefore no
787 return path is set.
78921. Internal re-arrangement: the function for sending a challenge and reading
790 a response while authentication was assuming a zero-terminated challenge
791 string. It's now changed to take a pointer and a length, to allow for
792 binary data in such strings.
79422. Added the cyrus_sasl authenticator (code supplied by MBM).
79623. Exim was not respecting finduser_retries when seeking the login of the
797 uid under which it was called; it was always trying 10 times. (The default
798 setting of finduser_retries is zero.) Also, it was sleeping after the final
799 failure, which is pointless.
80124. Implemented tls_on_connect_ports.
80325. Implemented acl_smtp_predata.
80526. If the domain in control=submission is set empty, Exim assumes that the
806 authenticated id is a complete email address when it generates From: or
807 Sender: header lines.
80927. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
810 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
811 chown and chgrp in /bin and hostname in /usr/bin.
81328. Exim was keeping the "process log" file open after each use, just as it
814 does for the main log. This opens the possibility of it remaining open for
815 long periods when the USR1 signal hits a daemon. Occasional processlog
816 errors were reported, that could have been caused by this. Anyway, it seems
817 much more sensible not to leave this file open at all, so that is what now
818 happens.
82029. The long-running daemon process does not normally write to the log once it
821 has entered its main loop, and it closes the log before doing so. This is
822 so that log files can straightforwardly be renamed and moved. However,
823 there are a couple of unusual error situations where the daemon does write
824 log entries, and I had neglected to close the log afterwards.
82630. The text of an SMTP error response that was received during a remote
827 delivery was being truncated at 512 bytes. This is too short for some of
828 the long messages that one sometimes sees. I've increased the limit to
829 1024.
83131. It is now possible to make retry rules that apply only when a message has a
832 specific sender, in particular, an empty sender.
83432. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
835 it possible to be selective about when SMTP synchronization is enforced.
83733. Added "control = caseful_local_part" and "control = "caselower_local_part".
83932. Implemented hosts_connection_nolog.
84133. Added an ACL for QUIT.
84334. Setting "delay_warning=" to disable warnings was not working; it gave a
844 syntax error.
84635. Added mailbox_size and mailbox_filecount to appendfile.
84836. Added control = no_multiline_responses to ACLs.
85037. There was a bug in the logic of the code that waits for the clock to tick
851 in the case where the clock went backwards by a substantial amount such
852 that the microsecond fraction of "now" was more than the microsecond
853 fraction of "then" (but the whole seconds number was less).
85538. Added support for the libradius Radius client library this is found on
856 FreeBSD (previously only the radiusclient library was supported).
859Exim version 4.42
862 1. When certain lookups returned multiple values in the form name=value, the
863 quoting of the values was not always being done properly. Specifically:
864 (a) If the value started with a double quote, but contained no whitespace,
865 it was not quoted.
866 (b) If the value contained whitespace other than a space character (i.e.
867 tabs or newlines or carriage returns) it was not quoted.
868 This fix has been applied to the mysql and pgsql lookups by writing a
869 separate quoting function and calling it from the lookup code. The fix
870 should probably also be applied to nisplus, ibase and oracle lookups, but
871 since I cannot test any of those, I have not disturbed their existing code.
873 2. A hit in the callout cache for a specific address caused a log line with no
874 reason for rejecting RCPT. Now it says "Previous (cached) callout
875 verification failure".
877 3. There was an off-by-one bug in the queryprogram router. An over-long
878 return line was truncated at 256 instead of 255 characters, thereby
879 overflowing its buffer with the terminating zero. As well as fixing this, I
880 have increased the buffer size to 1024 (and made a note to document this).
882 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
883 when Exim is waiting for an SMTP response from a remote server, Exim
884 restarts its select() call on the socket, thereby resetting its timeout.
885 This is not a problem when such interrupts are rare. Somebody set up a cron
886 job to run exiwhat every 2 minutes, which is less than the normal select()
887 timeout (5 or 10 minutes). This meant that the select() timeout never
888 kicked in because it was always reset. I have fixed this by comparing the
889 time when an interrupt arrives with the time at the start of the first call
890 to select(). If more time than the timeout has elapsed, the interrupt is
891 treated as a timeout.
893 5. Some internal re-factoring in preparation for the addition of Sieve
894 extensions (by MH). In particular, the "personal" test is moved to a
895 separate function, and given an option for scanning Cc: and Bcc: (which is
896 not set for Exim filters).
898 6. When Exim created an email address using the login of the caller as the
899 local part (e.g. when creating a From: or Sender: header line), it was not
900 quoting the local part when it contained special characters such as @.
902 7. Installed new OpenBSD configuration files.
904 8. Reworded some messages for syntax errors in "and" and "or" conditions to
905 try to make them clearer.
907 9. Callout options, other than the timeout value, were being ignored when
908 verifying sender addresses in header lines. For example, when using
910 verify = header_sender/callout=no_cache
912 the cache was (incorrectly) being used.
91410. Added a missing instance of ${EXE} to the exim_install script; this affects
915 only the Cygwin environment.
91711. When return_path_on_delivery was set as a log selector, if different remote
918 addresses in the same message used different return paths and parallel
919 remote delivery occurred, the wrong values would sometimes be logged.
920 (Whenever a remote delivery process finished, the return path value from
921 the most recently started remote delivery process was logged.)
92312. RFC 3848 specifies standard names for the "with" phrase in Received: header
924 lines when AUTH and/or TLS are in use. This is the "received protocol"
925 field. Exim used to use "asmtp" for authenticated SMTP, without any
926 indication (in the protocol name) for TLS use. Now it follows the RFC and
927 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
928 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
929 names appear in log lines as well as in Received: header lines.
93113. Installed MH's patches for Sieve to add the "copy" and "vacation"
932 extensions, and comparison tests, and to fix some bugs.
93414. Changes to the "personal" filter test:
936 (1) The test was buggy in that it was just doing the equivalent of
937 "contains" tests on header lines. For example, if a user's address was
938 anne@some.where, the "personal" test would incorrectly be true for
940 To: susanne@some.where
942 This test is now done by extracting each address from the header in turn,
943 and checking the entire address. Other tests that are part of "personal"
944 are now done using regular expressions (for example, to check local parts
945 of addresses in From: header lines).
947 (2) The list of non-personal local parts in From: addresses has been
948 extended to include "listserv", "majordomo", "*-request", and "owner-*",
949 taken from the Sieve specification recommendations.
951 (3) If the message contains any header line starting with "List-" it is
952 treated as non-personal.
954 (4) The test for "circular" in the Subject: header line has been removed
955 because it now seems ill-conceived.
95715. Minor typos in src/EDITME comments corrected.
95916. Installed latest exipick from John Jetmore.
96117. If headers_add on a router specified a text string that was too long for
962 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
963 of string_sprintf() is now avoided.
96518. $message_body_size was not set (it was always zero) when running the DATA
966 ACL and the local_scan() function.
96819. For the "mail" command in an Exim filter, no default was being set for
969 the once_repeat time, causing a random time value to be used if "once" was
970 specified. (If the value happened to be <= 0, no repeat happened.) The
971 default is now 0s, meaning "never repeat". The "vacation" command was OK
972 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
973 (I found it when inspecting the code).
97520. There is now an overall timeout for performing a callout verification. It
976 defaults to 4 times the callout timeout, which applies to individual SMTP
977 commands during the callout. The overall timeout applies when there is more
978 than one host that can be tried. The timeout is checked before trying the
979 next host. This prevents very long delays if there are a large number of
980 hosts and all are timing out (e.g. when the network connections are timing
981 out). The value of the overall timeout can be changed by specifying an
982 additional sub-option for "callout", called "maxwait". For example:
984 verify = sender/callout=5s,maxwait=20s
98621. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
987 to the end before writing, but this should make it even safer).
98922. Exim was forgetting that it had advertised PIPELINING for the second and
990 subsequent messages on an SMTP connection. It was also not resetting its
991 memory on STARTTLS and an internal HELO.
99323. When Exim logs an SMTP synchronization error within a session, it now
994 records whether PIPELINING has been advertised or not.
99624. Added 3 instances of "(long int)" casts to time_t variables that were being
997 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
998 rather than long int.
100025. Installed the latest Cygwin configuration files from the Cygwin maintainer.
100226. Added the never_mail option to autoreply.
1005Exim version 4.41
1008 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1009 crash if the getsockname() call failed; this can happen if a connection is
1010 closed very soon after it is established. The problem was simply in the
1011 order in which certain operations were done, causing Exim to try to write
1012 to the SMTP stream before it had set up the file descriptor. The bug has
1013 been fixed by making things happen in the correct order.
1016Exim version 4.40
1019 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1020 before the connection was closed, thus losing the rejection response.
1022 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1023 some early Solaris releases, but causes trouble in current releases where
1024 socklen_t is defined.
1026 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1027 always exist.
1029 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1030 configured.
1032 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1033 that releasing the top of it at the end releases what was used for sub-
1034 expansions (unless the block got too big). However, discard this block if
1035 the first thing is a variable or header, so that we can use its block when
1036 it is dynamic (useful for very large $message_headers, for example).
1038 6. Lookups now cache *every* query, not just the most recent. A new, separate
1039 store pool is used for this. It can be recovered when all lookup caches are
1040 flushed. Lookups now release memory at the end of their result strings.
1041 This has involved some general refactoring of the lookup sources.
1043 7. Some code has been added to the store_xxx() functions to reduce the amount
1044 of flapping under certain conditions.
1046 8. log_incoming_interface used to affect only the <= reception log lines. Now
1047 it causes the local interface and port to be added to several more SMTP log
1048 lines, for example "SMTP connection from", and rejection lines.
1050 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
105210. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
105411. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1055 could be overwritten at the end of the current message (or the start of a
1056 new message if it was set in a HELO ACL). The value is now preserved for
1057 the duration of the SMTP connection.
105912. If a transport had a headers_rewrite setting, and a matching header line
1060 contained an unqualified address, that address was qualified, even if it
1061 did not match any rewriting rules. The underlying bug was that the values
1062 of the flags that permit the existence of unqualified sender and recipient
1063 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1064 non-local messages, and by -bnq for local messages) were not being
1065 preserved with the message after it was received.
106713. When Exim was logging an SMTP synchronization error, it could sometimes log
1068 "next input=" as part of the text comprising the host identity instead of
1069 the correct text. The code was using the same buffer for two different
1070 strings. However, depending on which order the printing function evaluated
1071 its arguments, the bug did not always show up. Under Linux, for example, my
1072 test suite worked just fine.
107414. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1075 doesn't work with some older versions of Perl. It has been changed to "my",
1076 which in any case is probably the better facility to use.
107815. A really picky compiler found some instances of statements for creating
1079 error messages that either had too many or two few arguments for the format
1080 string.
108216. The size of the buffer for calls to the DNS resolver has been increased
1083 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1084 for addresses that have a lot of PTR records. This alleviates a problem; it
1085 does not fully solve it.
108717. A dnsdb lookup for PTR records that receives more data than will fit in the
1088 buffer now truncates the list and logs the incident, which is the same
1089 action as happens when Exim is looking up a host name and its aliases.
1090 Previously in this situation something unpredictable would happen;
1091 sometimes it was "internal error: store_reset failed".
109318. If a server dropped the connection unexpectedly when an Exim client was
1094 using GnuTLS and trying to read a response, the client delivery process
1095 crashed while trying to generate an error log message.
109719. If a "warn" verb in an ACL added multiple headers to a message in a single
1098 string, for example:
1100 warn message = H1: something\nH2: something
1102 the text was added as a single header line from Exim's point of view
1103 though it ended up OK in the delivered message. However, searching for the
1104 second and subsequent header lines using $h_h2: did not work. This has been
1105 fixed. Similarly, if a system filter added multiple headers in this way,
1106 the routers could not see them.
110820. Expanded the error message when iplsearch is called with an invalid key to
1109 suggest using net-iplsearch in a host list.
111121. When running tests using -bh, any delays imposed by "delay" modifiers in
1112 ACLs are no longer actually imposed (and a message to that effect is
1113 output).
111522. If a "gecos" field in a passwd entry contained escaped characters, in
1116 particular, if it contained a \" sequence, Exim got it wrong when building
1117 a From: or a Sender: header from that name. A second bug also caused
1118 incorrect handling when an unquoted " was present following a character
1119 that needed quoting.
112123. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1122 was not being matched caselessly.
112424. Arranged for all hyphens in the exim.8 source to be escaped with
1125 backslashes.
112725. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1128 properly. Recipient callout cache records were still being keyed to include
1129 the sender, even when use_sender was set false. This led to far more
1130 callouts that were necessary. The sender is no longer included in the key
1131 when use_sender is false.
113326. Added "control = submission" modifier to ACLs.
113527. Added the ${base62d: operator to decode base 62 numbers.
113728. dnsdb lookups can now access SRV records.
113929. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1140 the configuration file.
114230. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1143 (-v) mode. This makes the output for a verbose queue run more intelligible.
114531. Added a use_postmaster feature to recipient callouts.
114732. Added the $body_zerocount variable, containing the number of binary zero
1148 bytes in the message body.
115033. The time of last modification of the "new" subdirectory is now used as the
1151 "mailbox time last read" when there is a quota error for a maildir
1152 delivery.
115434. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
115635. Added +ignore_unknown as a special item in host lists.
115836. Code for decoding IPv6 addresses in host lists is now included, even if
1159 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1160 address was recognized as an IP address, but was then not correctly decoded
1161 into binary, causing unexpected and incorrect effects when compared with
1162 another IP address.
1165Exim version 4.34
1168 1. Very minor rewording of debugging text in manualroute to say "list of
1169 hosts" instead of "hostlist".
1171 2. If verify=header_syntax was set, and a header line with an unqualified
1172 address (no domain) and a large number of spaces between the end of the
1173 name and the colon was received, the reception process suffered a buffer
1174 overflow, and (when I tested it) crashed. This was caused by some obsolete
1175 code that should have been removed. The fix is to remove it!
1177 3. When running in the test harness, delay a bit after writing a bounce
1178 message to get a bit more predictability in the log output.
1180 4. Added a call to search_tidyup() just before forking a reception process. In
1181 theory, someone could use a lookup in the expansion of smtp_accept_max_
1182 per_host which, without the tidyup, could leave open a database connection.
1184 5. Added the variables $recipient_data and $sender_data which get set from a
1185 lookup success in an ACL "recipients" or "senders" condition, or a router
1186 "senders" option, similar to $domain_data and $local_part_data.
1188 6. Moved the writing of debug_print from before to after the "senders" test
1189 for routers.
1191 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1192 problems for message scanning, either using a data ACL, or using
1193 local_scan() because the Received: header was not generated till after they
1194 were called (in order to set the time as the time of reception completion).
1195 I have revised the way this works. The header is now generated after the
1196 body is received, but before the ACL or local_scan() are called. After they
1197 are run, the timestamp in the header is updated.
1200Exim version 4.33
1203 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1204 before starting a queue runner without re-exec. This happened only when
1205 deliver_drop_privilege was set or when the Exim user was set to root. The
1206 effect of the bug was that timeouts during subsequent deliveries caused
1207 crashes instead of being properly handled. The handler is now left at its
1208 default (and expected) setting.
1210 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1211 message, again when deliver_drop_privilege is set or Exim is run as root.
1212 The bug described in (1) was not present in this case, but the tidying up
1213 of the other signals was missing. I have made the two cases consistent.
1215 3. The ignore_target_hosts setting on a manualroute router was being ignored
1216 for hosts that were looked up using the /MX notation.
1218 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1219 in domain lists.
1221 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1222 operated on the sender address. After changing the $sender_address to <>
1223 for the sender address verify, Exim was re-instated it as the original
1224 (before rewriting) address, but remembering that it had rewritten it, so it
1225 wasn't rewriting it again. This bug also had the effect of breaking the
1226 sender address verification caching when the sender address was rewritten.
1228 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1229 This has been changed so that if the ip literal address matches
1230 ignore_target_hosts, the router declines.
1232 7. Added expansion conditions match_domain, match_address, and match_local_
1233 part (NOT match_host).
1235 8. The placeholder for the Received: header didn't have a length field set.
1237 9. Added code to Exim itself and to exim_lock to test for a specific race
1238 condition that could lead to file corruption when using MBX delivery. The
1239 issue is with the lockfile that is created in /tmp. If this file is removed
1240 after a process has opened it but before that process has acquired a lock,
1241 there is the potential for a second process to recreate the file and also
1242 acquire a lock. This could lead to two Exim processes writing to the file
1243 at the same time. The added code performs the same test as UW imapd; it
1244 checks after acquiring the lock that its file descriptor still refers to
1245 the same named file.
124710. The buffer for building added header lines was of fixed size, 8192 bytes.
1248 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1249 when Exim is built.
125111. Added the smtp_active_hostname option. If used, this will typically be made
1252 to depend on the incoming interface address. Because $interface_address is
1253 not set up until the daemon has forked a reception process, error responses
1254 that can happen earlier (such as "too many connections") no longer contain
1255 a host name.
125712. If an expansion in a condition on a "warn" statement fails because a lookup
1258 defers, the "warn" statement is abandoned, and the next ACL statement is
1259 processed. Previously this caused the whole ACL to be aborted.
126113. Added the iplsearch lookup type.
126314. Added ident_timeout as a log selector.
126515. Added tls_certificate_verified as a log selector.
126716. Added a global option tls_require_ciphers (compare the smtp transport
1268 option of the same name). This controls incoming TLS connections.
127017. I finally figured out how to make tls_require_ciphers do a similar thing
1271 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1272 before starting the TLS session.
127418. Tabs are now shown as \t in -bP output.
127619. If the log selector return_path_on_delivery was set, Exim crashed when
1277 bouncing a message because it had too many Received: header lines.
127920. If two routers both had headers_remove settings, and the first one included
1280 a superfluous trailing colon, the final name in the first list and the
1281 first name in the second list were incorrectly joined into one item (with a
1282 colon in the middle).
1285Exim version 4.32
1288 1. Added -C and -D options to the exinext utility, mainly to make it easier
1289 to include in the automated testing, but these could be helpful when
1290 multiple configurations are in use.
1292 2. The exinext utility was not formatting the output nicely when there was
1293 an alternate port involved in the retry record key, nor when there was a
1294 message id as well (for retries that were specific to a specific message
1295 and a specific host). It was also confused by IPv6 addresses, because of
1296 the additional colons they contain. I have fixed the IPv4 problem, and
1297 patched it up to do a reasonable job for IPv6.
1299 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1300 delivery, the log line now contains "pipelined" if PIPELINING was used.
1302 4. An SMTP transport process used to panic and die if the bind() call to set
1303 an explicit outgoing interface failed. This has been changed; it is now
1304 treated in the same way as a connect() failure.
1306 5. A reference to $sender_host_name in the part of a conditional expansion
1307 that was being skipped was still causing a DNS lookup. This no longer
1308 occurs.
1310 6. The def: expansion condition was not recognizing references to header lines
1311 that used bh_ and bheader_.
1313 7. Added the _cache feature to named lists.
1315 8. The code for checking quota_filecount in the appendfile transport was
1316 allowing one more file than it should have been.
1318 9. For compatibility with Sendmail, the command line option
1320 -prval:sval
1322 is equivalent to
1324 -oMr rval -oMs sval
1326 and sets the incoming protocol and host name (for trusted callers). The
1327 host name and its colon can be omitted when only the protocol is to be set.
1328 Note the Exim already has two private options, -pd and -ps, that refer to
1329 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1330 "s", but I don't think that's a major issue.
133210. A number of refactoring changes to the code, none of which should affect
1333 Exim's behaviour:
1335 (a) The number of logging options was getting close to filling up the
1336 32-bit word that was used as a bit map. I have split them into two classes:
1337 those that are passed in the argument to log_write(), and those that are
1338 only ever tested independently outside of that function. These are now in
1339 separate 32-bit words, so there is plenty of room for expansion again.
1340 There is no change in the user interface or the logging behaviour.
1342 (b) When building, for example, log lines, the code previously used a
1343 macro that called string_cat() twice, in order to add two strings. This is
1344 not really sufficiently general. Furthermore, there was one instance where
1345 it was actually wrong because one of the argument was used twice, and in
1346 one call a function was used. (As it happened, calling the function twice
1347 did not affect the overall behaviour.) The macro has been replaced by a
1348 function that can join an arbitrary number of extra strings onto a growing
1349 string.
1351 (c) The code for expansion conditions now uses a table and a binary chop
1352 instead of a serial search (which was left over from when there were very
1353 few conditions). Also, it now recognizes conditions like "pam" even when
1354 the relevant support is not compiled in: a suitably worded error message is
1355 given if an attempt is made to use such a condition.
135711. Added ${time_interval:xxxxx}.
135912. A bug was causing one of the ddress fields not to be passed back correctly
1360 from remote delivery subprocesses. The field in question was not being
1361 subsequently used, so this caused to problems in practice.
136313. Added new log selectors queue_time and deliver_time.
136514. Might have fixed a bug in maildirsizefile handling that threw up
1366 "unexpected character" debug warnings, and recalculated the data
1367 unnecessarily. In any case, I expanded the warning message to give more
1368 information.
137015. Added the message "Restricted characters in address" to the statements in
1371 the default ACL that block characters like @ and % in local parts.
137316. Change 71 for release 4.31 proved to be much less benign that I imagined.
1374 Three changes have been made:
1376 (a) There was a serious bug; a negative response to MAIL caused the whole
1377 recipient domain to be cached as invalid, thereby blocking all messages
1378 to all local parts at the same domain, from all senders. This bug has
1379 been fixed. The domain is no longer cached after a negative response to
1380 MAIL if the sender used is not empty.
1382 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1383 been restored.
1385 (c) A new callout option, "use_sender" has been added for people who want
1386 the modified behaviour.
1389Exim version 4.31
1392 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1393 Larry Rosenman.
1395 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1396 indeed breaks things for older releases.
1398 3. Added additional logging to the case where there is a problem reading data
1399 from a filter that is running in a subprocess using a pipe, in order to
1400 try to track down a specific problem.
1402 4. Testing facility fudge: when running in the test harness and attempting
1403 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1404 getting "No route to host". Convert this to a timeout.
1406 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1407 warning.
1409 6. Some OS don't have socklen_t but use size_t instead. This affects the
1410 fifth argument of getsockopt() amongst other things. This is now
1411 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1412 can be set for individual OS. I have set it for SunOS5, OSF1, and
1413 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1414 some earlier ones do not.
1416 7. Change 4.30/15 was not doing the test caselessly.
1418 8. The standard form for an IPv6 address literal was being rejected by address
1419 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1420 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1421 this, as well as the form without the "IPv6" on the front (but only when
1422 address literals are enabled, of course).
1424 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
142610. Exim crashed if a message with an empty sender address specified by -f
1427 encountered a router with an errors_to setting. This could be provoked only
1428 by a command such as
1430 exim -f "" ...
1432 where an empty string was supplied; "<>" did not hit this bug.
143411. Installed PCRE release 4.5.
143612. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1437 remained set. It is now erased.
143913. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1440 times from message ids (which are base 36 rather than the normal 62).
144214. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1443 were being counted as actual protocol errors, and logged if the log
1444 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1445 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1446 and DATA following a set of rejected RCPTs do not count as protocol errors.
1447 In other words, Exim assumes they were pipelined, though this may not
1448 actually be the case. Of course, in all cases the client gets an
1449 appropriate error code.
145115. If a lookup fails in an ACL condition, a message about the failure may
1452 be available; it is used if testing the ACL cannot continue, because most
1453 such messages specify what the cause of the deferral is. However, some
1454 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1455 that caused an old message to be retained and used if a later statement
1456 caused a defer, replacing the real cause of the deferral.
145816. If an IP address had so many PTR records that the DNS lookup buffer
1459 was not large enough to hold them, Exim could crash while trying to process
1460 the truncated data. It now detects and logs this case.
146217. Further to 4.21/58, another change has been made: if (and only if) the
1463 first line of a message (the first header line) ends with CRLF, a bare LF
1464 in a subsequent header line has a space inserted after it, so as not to
1465 terminate the header.
146718. Refactoring: tidied an ugly bit of code in appendfile that copied data
1468 unnecessarily, used atoi() instead of strtol(), and didn't check the
1469 termination when getting file sizes from file names by regex.
147119. Completely re-implemented the support for maildirsize files, in the light
1472 of a number of problems with the previous contributed implementation
1473 (4.30/29). In particular:
1475 . If the quota is zero, the maildirsize file is maintained, but no quota is
1476 imposed.
1478 . If the maildir directory does not exist, it is created before any attempt
1479 to write a maildirsize file.
1481 . The quota value in the file is just a cache; if the quota is changed in
1482 the transport, the new value overrides.
1484 . A regular expression is available for excluding directories from the
1485 count.
148720. The autoreply transport checks the characters in options that define the
1488 message's headers; it allows continued headers, but it was checking with
1489 isspace() after an embedded newline instead of explicitly looking for a
1490 space or a tab.
149221. If all the "regular" hosts to which an address was routed had passed their
1493 expiry times, and had not reached their retry times, the address was
1494 bounced, even if fallback hosts were defined. Now Exim should go on to try
1495 the fallback hosts.
149722. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1498 equivalent code in the SMTP transport. Some hosts send humungous responses
1499 to HELO/EHLO, more than 1024 it seems.
150123. Refactoring: code in filter.c used (void *) for "any old type" but this
1502 gives compiler warnings in some environments. I've now done it "properly",
1503 using a union.
150524. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1506 (because of problems with the built-in one) was declared to return uschar *
1507 instead of char *, causing compiler failure.
150925. Fixed a file descriptor leak when processing alias/forward files.
151126. Fixed a minor format string issue in dbfn.c.
151327. Typo in exim.c: ("dmbnz" for "dbmnz").
151528. If a filter file refered to $h_xxx or $message_headers, and the headers
1516 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1517 become corrupted.
151929. When a sender address is verified, it is cached, to save repeating the test
1520 when there is more than one recipient in a message. However, when the
1521 verification involves a callout, it is possible for different callout
1522 options to be set for different recipients. It is too complicated to keep
1523 track of this in the cache, so now Exim always runs a verification when a
1524 callout is required, relying on the callout cache for the optimization.
1525 The overhead is duplication of the address routing, but this should not be
1526 too great.
152830. Fixed a bug in callout caching. If a RCPT command caused the sender address
1529 to be verified with callout=postmaster, and the main callout worked but the
1530 postmaster check failed, the verification correctly failed. However, if a
1531 subsequent RCPT command asked for sender verification *without* the
1532 postmaster check, incorrect caching caused this verification also to fail,
1533 incorrectly.
153531. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1536 it was not caching the DNS options (qualify_single, search_parents) that
1537 were used when the lookup failed. A subsequent lookup with different
1538 options therefore always gave the same answer, though there were cases
1539 where it should not have. (Example: a "domains = !$mx_any" option on a
1540 dnslookup router: the "domains" option is always processed without any
1541 widening, but the router might have qualify_single set.) Now Exim uses the
1542 cached value only when the same options are set.
154432. Added John Jetmore's "exipick" utility to the distribution.
154633. GnuTLS: When an attempt to start a TLS session fails for any reason other
1547 than a timeout (e.g. a certificate is required, and is not provided), an
1548 Exim server now closes the connection immediately. Previously it waited for
1549 the client to close - but if the client is SSL, it seems that they each
1550 wait for each other, leading to a delay before one of them times out.
155234: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1553 maintained 0.8.x compatibility because I don't think many are using it, and
1554 it is clearly obsolete.
155635. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1557 transport.
155936. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1560 client certificate was expired. A simple patch fixes this, though I don't
1561 understand the full logic of why the verify callback is called multiple
1562 times.
156437. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1565 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1566 which causes problems with some clients (such as the Certicom SSL Plus
1567 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1568 disables the coutermeasure allowing Eudora to connect."
157038. Exim was not checking that a write() to a log file succeeded. This could
1571 lead to Bad Things if a log got too big, in particular if it hit a file
1572 size limit. Exim now panics and dies if it cannot write to a log file, just
1573 as it does if it cannot open a log file.
157539. Modified OS/Makefile-Linux so that it now contains
1579 The two -D definitions ensure that Exim is compiled with large file
1580 support, which makes it possible to handle log files that are bigger than
1581 2^31.
158340. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1584 instance) a domain was checked against a named list that involved a lookup,
1585 causing $domain_data to be set, then another domain was checked against the
1586 same list, then the first domain was re-checked, the value of $domain_data
1587 after the final check could be wrong. In particular, if the second check
1588 failed, it could be set empty. This bug probably also applied to
1589 $localpart_data.
159141. The strip_trailing_dot option was not being applied to the address given
1592 with the -f command-line option.
159442. The code for reading a message's header from the spool was incrementing
1595 $received_count, but never initializing it. This meant that the value was
1596 incorrect (doubled) while delivering a message in the same process in which
1597 it was received. In the most common configuration of Exim, this never
1598 happens - a fresh exec is done - but it can happen when
1599 deliver_drop_privilege is set.
160143. When Exim logs an SMTP synchronization error - client data sent too soon -
1602 it now includes up to 150 characters of the unexpected data in the log
1603 line.
160544. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1606 and building data strings. The size of both of these buffers was 10 000
1607 bytes - far larger than anybody would *ever* want, thought I. Needless to
1608 say, somebody hit the limit. I have increased the maximum line length to
1609 20 000 and the maximum data length of concatenated lines to 100 000. I have
1610 also fixed two bugs, because there was no checking on these buffers. Tsk,
1611 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1612 buffer is too small.
161445. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1615 lsearch lookups. Now it does.
161746. When parsing a route_list item in a manualroute router, a fixed-length
1618 buffer was used for the list of hosts. I made this 1024 bytes long,
1619 thinking that nobody would ever have a list of hosts that long. Wrong.
1620 Somebody had a whole pile of complicated expansion conditions, and the
1621 string was silently truncated, leading to an expansion error. It turns out
1622 that it is easier to change to an unlimited length (owing to other changes
1623 that have happened since this code was originally written) than to build
1624 structure for giving a limitation error. The length of the item that
1625 expands into the list of hosts is now unlimited.
162747. The lsearch lookup could not handle data where the length of text line was
1628 more than 4095 characters. Such lines were truncated, leading to shortened
1629 data being returned. It should now handle lines of any length.
163148. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1632 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1633 ACL").
163549. Cosmetic tidy to scripts like exicyclog that are generated by globally
1636 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1637 no longer happens in comment lines. A list of replacements is now placed
1638 at the head of all of the source files, except those whose only change is
1639 to replace PERL_COMMAND in the very first #! line.
164150. Replaced the slow insertion sort in queue.c, for sorting the list of
1642 messages on the queue, with a bottom-up merge sort, using code contributed
1643 by Michael Haardt. This should make operations like -bp somewhat faster on
1644 large queues. It won't affect queue runners, except when queue_run_in_order
1645 is set.
164751. Installed eximstats 1.31 in the distribution.
164952. Added support for SRV lookups to the dnslookup router.
165153. If an ACL referred to $message_body or $message_body_end, the value was not
1652 reset for any messages that followed in the same SMTP session.
165454. The store-handling optimization for building very long strings was not
1655 differentiating between the different store pools. I don't think this
1656 actually made any difference in practice, but I've tidied it.
165855. While running the routers to verify a sender address, $sender_address
1659 was still set to the sender address. This is wrong, because when routing to
1660 send a bounce to the sender, it would be empty. Therefore, I have changed
1661 it so that, while verifying a sender address, $sender_address is set to <>.
1662 (There is no change to what happens when verifying a recipient address.)
166456. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1665 target A or AAAA records (if not already returned) without resetting the
1666 qualify_single or search_parents options of the DNS resolver. These are
1667 inappropriate in this case because the targets of MX and SRV records must
1668 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1669 target that, when qualified, matched something in the local domain. These
1670 two options are now turned off when doing these lookups.
167257. It seems that at least some releases of Reiserfs (which does not have the
1673 concept of a fixed number of inodes) returns zero and not -1 for the
1674 number of available inodes. This interacted badly with check_spool_inodes,
1675 which assumed that -1 was the "no such thing" setting. What I have done is
1676 to check that the total number of inodes is greater than zero before doing
1677 the test of how many are available.
167958. When a "warn" ACL statement has a log_message modifier, the message is
1680 remembered, and not repeated. This is to avoid a lot of repetition when a
1681 message has many recipients that cause the same warning to be written.
1682 Howewer, Exim was preserving the list of already written lines for an
1683 entire SMTP session, which doesn't seem right. The memory is now reset if a
1684 new message is started.
168659. The "rewrite" debugging flag was not showing the result of rewriting in the
1687 debugging output unless log_rewrite was also set.
168960. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1690 of (int)(handle) when we know that handle contains (void *)(-1).
169261. The Exim daemon panic-logs an error return when it closes the incoming
1693 connection. However "connection reset by peer" seems to be common, and
1694 isn't really an error worthy of noting specially, so that particular error
1695 is no long logged.
169762. When Exim is trying to find all the local interfaces, it used to panic and
1698 die if the ioctl to get the interface flags failed. However, it seems that
1699 on at least one OS (Solaris 9) it is possible to have an interface that is
1700 included in the list of interfaces, but for which you get a failure error
1701 for this call. This happens when the interface is not "plumbed" into a
1702 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1703 failure of the "get flags" call assumes that the interface is down.
170563. Added a ${eval10: operator, which assumes all numbers are decimal. This
1706 makes life easier for people who are doing arithmetic on fields extracted
1707 from dates, where you often get leading zeros that should not be
1708 interpreted as octal.
171064. Added qualify_domain to the redirect router, to override the global
1711 setting.
171365. If a pathologically long header line contained very many addresses (the
1714 report of this problem mentioned 10 000) and each of them was rewritten,
1715 Exim could use up a very large amount of memory. (It kept on making new
1716 copies of the header line as it rewrote, and never released the old ones.)
1717 At the expense of a bit more processing, the header rewriting function has
1718 been changed so that it no longer eats memory in this way.
172066. The generation of the Received: header has been moved from the time that a
1721 message starts to be received, to the time that it finishes. The timestamp
1722 in the Received: header should now be very close to that of the <= log
1723 line. There are two side-effects of this change:
1725 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1726 logged header lines no longer include the local Received: line, because
1727 it has not yet been created. The same applies to a copy of the message
1728 that is returned to a non-SMTP sender when a message is rejected.
1730 (b) When a filter file is tested using -bf, no additional Received: header
1731 is added to the test message. After some thought, I decided that this
1732 is a bug fix.
1734 This change does not affect the value of $received_for. It is still set
1735 after address rewriting, but before local_scan() is called.
173767. Installed the latest Cygwin-specific files from the Cygwin maintainer.
173968. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1740 gave an unhelpful panic error message, and a defer error. I have managed to
1741 change this behaviour so that it now rejects any supplied certificate,
1742 which seems right, as the list of acceptable certificates is empty.
174469. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1745 gave an unhelpful defer error. I have not managed to make this reject any
1746 supplied certificates, but the error message it gives is "no certificate
1747 supplied", which is not helpful.
174970. exigrep's output now also includes lines that are not associated with any
1750 message, but which match the given pattern. Implemented by a patch from
1751 Martin Sluka, which also tidied up the Perl a bit.
175371. Recipient callout verification, like sender verification, was using <> in
1754 the MAIL FROM command. This isn't really the right thing, since the actual
1755 sender may affect whether the remote host accepts the recipient or not. I
1756 have changed it to use the actual sender in the callout; this means that
1757 the cache record is now keyed on a recipient/sender pair, not just the
1758 recipient address. There doesn't seem to be a real danger of callout loops,
1759 since a callout by the remote host to check the sender would use <>.
1760 [SEE ABOVE: changed after hitting problems.]
176272. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1763 temporary errors. However, in the case of such a code being given after
1764 the end of a data transmission (i.e. after ".") Exim was failing to write
1765 a retry record for the message. (Yes, there was some broken host that was
1766 actually sending 8xx at this point.)
176873. An unknown lookup type in a host list could cause Exim to panic-die when
1769 the list was checked. (An example that provoked this was putting <; in the
1770 middle of a list instead of at the start.) If this happened during a DATA
1771 ACL check, a -D file could be left lying around. This kind of configuration
1772 error no longer causes Exim to die; instead it causes a defer errror. The
1773 incident is still logged to the main and panic logs.
177574. Buglet left over from Exim 3 conversion. The message "too many messages
1776 in one connection" was written to the rejectlog but not the mainlog, except
1777 when address rewriting (yes!) was being logged.
177975. Added write_rejectlog option.
178176. When a system filter was run not as root (that is, when system_filter_user
1782 was set), the values of the $n variables were not being returned to the
1783 main process; thus, they were not subsequently available in the $sn
1784 variables.
178677. Added +return_path_on_delivery log selector.
178878. A connection timeout was being treated differently from recipients deferred
1789 when testing hosts_max_try with a message that was older than the host's
1790 retry timeout. (The host should not be counted, thus allowing all hosts to
1791 be tried at least once before bouncing.) This may have been the cause of an
1792 occasionally reported bug whereby a message would remain on the queue
1793 longer than the retry timeout, but would be bounced if a delivery was
1794 forced. I say "may" because I never totally pinned down the problem;
1795 setting up timeout/retry tests is difficult. See also the next item.
179779. The ultimate address timeout was not being applied to errors that involved
1798 a combination of host plus message (for example, a timeout on a MAIL
1799 command). When an address resolved to a number of possible hosts, and they
1800 were not all tried for each delivery (e.g. because of hosts_max_try), a
1801 message could remain on the queue longer than the retry timeout.
180380. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1804 Haardt.
180681. Fixed an obscure SMTP outgoing bug which required at least the following
1807 conditions: (a) there was another message waiting for the same server;
1808 (b) the server returned 5xx to all RCPT commands in the first message so
1809 that the message was not completed; (c) the server dropped the connection
1810 or gave a negative response to the RSET that Exim sends to abort the
1811 transaction. The observed case was a dropped connection after DATA that had
1812 been sent in pipelining mode. That is, the server had advertised PIPELINING
1813 but was not implementing it correctly. The effect of the bug was incorrect
1814 behaviour, such as trying another host, and this could lead to a crash.
1817Exim version 4.30
1820 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1821 and daemon.c were passed as pointers to ints; they should have been
1822 pointers to socklen_t variables (which are typically unsigned ints).
1824 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1825 fixed.
1827 3. Fixed a really odd bug that affected only the testing scheme; patching a
1828 certain fixed string in the binary changed the value of another string that
1829 happened to be identical to the end of the original first string.
1831 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1832 name", it returns that address as the IP address. On some operating
1833 systems (e.g. Solaris), it also passes back the IP address string as the
1834 "host name". However, on others (e.g. Linux), it passes back an empty
1835 string. Exim wasn't checking for this, and was changing the host name to an
1836 empty string, assuming it had been canonicized.
1838 5. Although rare, it is permitted to have more than one PTR record for a given
1839 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1840 all the names associated with an address, because they do in Solaris.
1841 However, it seems that they do not in Linux for data that comes from the
1842 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1843 I found this out when I moved to a new Linux workstation and tried to run
1844 the Exim test suite.
1846 To get round this problem I have changed the code so that it now does its
1847 own call to the DNS to look up PTR records when searching for a host name.
1848 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1849 addresses that are only in /etc/hosts are still found.
1851 This behaviour is, however, controlled by an option called host_lookup_
1852 order, which defaults to "bydns:byaddr". If people want to use the other
1853 order, or indeed, just use one or the other means of lookup, they can
1854 specify it in this variable.
1856 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1857 some operating systems, this comes back from gethostbyaddr() as an empty
1858 string, and this is what Exim used to test for. However, it seems that in
1859 other systems, "." is yielded. Exim now tests for this case too.
1861 7. The values of check_spool_space and check_log_space are now held internally
1862 as a number of kilobytes instead of an absolute number of bytes. If a
1863 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1864 kilobyte. This means that much larger values can be stored.
1866 8. Exim monitor: an attempt to get the action menu when not actually pointing
1867 at a message produces an empty menu entitled "No message selected". This
1868 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1869 no entries in it ("Shell widget menu has zero width and/or height"). So I
1870 have added a single, blank menu entry in this case.
1872 9. Added ${quote_local_part.
187410. MIME decoding is now applied to the contents of Subject: header lines when
1875 they are logged.
187711. Now that a reference to $sender_host_address automatically causes a reverse
1878 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1879 host lookup before query-style lookups in lists that might use this
1880 variable. This has therefore been abolished, and the "net-" prefix is no
1881 longer necessary for query-style lookups.
188312. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1884 have been a typo for LFLAGS, so it has been changed.
188613. The install script calls Exim with "-C /dev/null" in order to find the
1887 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1888 to be output. Howeve, since Exim outputs its version number before the
1889 error, it didn't break the script. It just looked ugly. I fixed this by
1890 always allowing "-C /dev/null" if the caller is root.
189214. Ignore overlarge ACL variable number when reading spool file - insurance
1893 against a later release with more variables having written the file.
189515. The standard form for an IPv6 address literal was being rejected by EHLO.
1896 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1897 this, as well as the form without the "IPv6" on the front.
189916. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1900 OS/Makefile-Darwin file.
190217. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1903 only to LDAP libraries that do not have LDAP_OPT_DEREF.
190518. After change 4.21/52, "%ld" was used to format the contents of the $inode
1906 variable. However, some OS use ints for inodes. I've added cast to long int
1907 to get rid of the compiler warning.
190919. I had forgotten to lock out "/../" in configuration file names when
1910 ALT_CONFIG_PREFIX was set.
191220. Routers used for verification do not need to specify transports. However,
1913 if such a router generated a host list, and callout was configured, Exim
1914 crashed, because it could not find a port number from the (non-existent)
1915 transport. It now assumes port 25 in this circumstance.
191721. Added the -t option to exigrep.
191922. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1920 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1921 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1922 them is set, LOOKUP_LSEARCH is forced.
192423. "exim -bV" now outputs a list of lookups that are included in the binary.
192624. Added sender and host information to the "rejected by local_scan()" log
1927 line; previously there was no indication of these.
192925. Added .include_if_exists.
193126. Change 3.952/11 added an explicit directory sync on top of a file sync for
1932 Linux. It turns out that not all file systems support this. Apparently some
1933 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1934 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1935 not supported on the file descriptor, is now ignored when Exim is trying to
1936 sync a directory. This applies only to Linux.
193827. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
194028. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1941 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1942 this.
194429. Added support for maildirsize files from supplied patch (modified a bit).
194630. The use of :fail: followed by an empty string could lead Exim to respond to
1947 sender verification failures with (e.g.):
1949 550 Verification failed for <xxx>
1950 550 Sender verify failed
1952 where the first response line was missing the '-' that indicates it is not
1953 the final line of the response.
195531. The loop for finding the name of the user that called Exim had a hardwired
1956 limit of 10; it now uses the value of finduser_retries, which is used for
1957 all other user lookups.
195932. Added $received_count variable, available in data and not_smtp ACLs, and at
1960 delivery time.
196233. Exim was neglecting to zero errno before one call of strtol() when
1963 expanding a string and expecting an integer value. On some systems this
1964 resulted in spurious "integer overflow" errors. Also, it was casting the
1965 result into an int without checking.
196734. Testing for a connection timeout using "timeout_connect" in the retry rules
1968 did not work. The code looks as if it has *never* worked, though it appears
1969 to have been documented since at least releast 1.62. I have made it work.
197135. The "timeout_DNS" error in retry rules, also documented since at least
1972 1.62, also never worked. As it isn't clear exactly what this means, and
1973 clearly it isn't a major issue, I have abolished the feature by treating it
1974 as "timeout", and writing a warning to the main and panic logs.
197636. The display of retry rules for -brt wasn't always showing the error code
1977 correctly.
197937. Added new error conditions to retry rules: timeout_A, timeout_MX,
1980 timeout_connect_A, timeout_connect_MX.
198238. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1983 to the empty sender.
198539. The daemon was not analysing the content of -oX till after it had closed
1986 stderr and disconnected from the controlling terminal. This meant that any
1987 syntax errors were only noted on the panic log, and the return code from
1988 the command was 0. By re-arranging the code a little, I've made the
1989 decoding happen first, so such errors now appear on stderr, and the return
1990 code is 1. However, the actual setting up of the sockets still happens in
1991 the disconnected process, so errors there are still only recorded on the
1992 panic log.
199440. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1995 connections (as happens on some IP stacks) was logged at start up time as
1996 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1997 it from "IPv6 and IPv4", which means that two separate sockets are being
1998 used.
200041. The debug output for gethostbyname2() or getipnodebyname() failures now
2001 says whether AF_INET or AF_INET6 was passed as an argument.
200342. Exiwhat output was messed up when time zones were included in log
2004 timestamps.
200643. Exiwhat now gives more information about the daemon's listening ports,
2007 and whether -tls-on-connect was used.
200944. The "port" option of the smtp transport is now expanded.
201145. A "message" modifier in a "warn" statement in a non-message ACL was being
2012 silently ignored. Now an error message is written to the main and panic
2013 logs.
201546. There's a new ACL modifier called "logwrite" which writes to a log file
2016 as soon as it is encountered.
201847. Added $local_user_uid and $local_user_gid at routing time.
202048. Exim crashed when trying to verify a sender address that was being
2021 rewritten to "<>".
202349. Exim was recognizing only a space character after ".include". It now also
2024 recognizes a tab character.
202650. Fixed several bugs in the Perl script that creates the exim.8 man page by
2027 extracting the relevant information from the specification. The man page no
2028 longer contains scrambled data for the -d option, and I've added a section
2029 at the front about calling Exim under different names.
203151. Added "extra_headers" argument to the "mail" command in filter files.
203352. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2034 crash.
203653. Installed eximstats 1.29.
203854. Added transport_filter_timeout as a generic transport option.
204055. Exim no longer adds an empty Bcc: header to messages that have no To: or
2041 Cc: header lines. This was required by RFC 822, but it not required by RFC
2042 2822.
204456. Exim used to add From:, Date:, and Message-Id: header lines to any
2045 incoming messages that did not have them. Now it does so only if the
2046 message originates locally, that is, if there is no associated remote host
2047 address. When Resent- header lines are present, this applies to the Resent-
2048 lines rather than the non-Resent- lines.
205057. Drop incoming SMTP connection after too many syntax or protocol errors. The
2051 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
205358. Messages for configuration errors now include the name of the main
2054 configuration file - useful now that there may be more than one file in a
2055 list (.included file names were always shown).
205759. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2058 for those rare installations that do not start the daemon as root or run it
2059 setuid root. I've cut out the call to initgroups() if the daemon is not
2060 root at that time.
206260. The Exim user and group can now be bound into the binary as text strings
2063 that are looked up at the start of Exim's processing.
206561. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
206762. Added $mailstore_basename variable.
206963. Installed patch to sieve.c from Michael Haardt.
207164. When Exim failed to open the panic log after failing to open the main log,
2072 the original message it was trying to log was written to stderr and debug
2073 output, but if they were not available (the usual case in production), it
2074 was lost. Now it is written to syslog before the two lines that record the
2075 failures to open the logs.
207765. Users' Exim filters run in subprocesses under the user's uid. It is
2078 possible for a "deliver" command or an alias in a "personal" command to
2079 provoke an address rewrite. If logging of address rewriting is configured,
2080 this fails because the process is not running as root or exim. There may be
2081 a better way of dealing with this, but for the moment (because 4.30 needs
2082 to be released), I have disabled address rewrite logging when running a
2083 filter in a non-root, non-exim process.
2086Exim version 4.24
2089 1. The buildconfig auxiliary program wasn't quoting the value set for
2090 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2091 not defined. This bug was masked in 4.22 by the effect that was fixed in
2092 change 4.23/1.
2094 2. Some messages that were rejected after a message id was allocated were
2095 shown as "incomplete" by exigrep. It no longer does this for messages that
2096 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2098 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2099 have allow_domain_literals set, the ID did not get logged in the <= line.
2100 Domain literals are now always recognized in Message-ID: header lines.
2102 4. The first argument for a ${extract expansion item is the key name or field
2103 number. Leading and trailing spaces in this item were not being ignored,
2104 causing some misleading effects.
2106 5. When deliver_drop_privilege was set, single queue runner processes started
2107 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2108 same command in the process it spins off) were not dropping privilege.
2110 6. When the daemon running as "exim" started a queue runner, it always
2111 re-executed Exim in the spun-off process. This is a waste of effort when
2112 deliver_drop_privilege is set. The new process now just calls the
2113 queue-runner function directly.
2116Exim version 4.23
2119 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2122 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2123 ignored. Though the use of -oP was forcing the writing of a pid file, it
2124 was always written to the default place.
2126 3. If the message "no IP address found for host xxxx" is generated during
2127 incoming verification, it is now followed by identification of the incoming
2128 connection (so you can more easily find what provoked it).
2130 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2132 5. Added some features to "harden" Exim a bit more against certain attacks:
2134 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2135 be put in Local/Makefile. This is like the never_users runtime option,
2136 but it cannot be overridden. The default setting is "root".
2138 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2139 prefix string with which any file named in a -C command line option
2140 must start.
2142 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2143 is retained for -C and -D only if the caller of Exim is root. Without
2144 it, the exim user may also use -C and -D and retain privilege.
2146 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2147 command line option is disabled.
2149 6. Macro names set by the -D option must start with an upper case letter, just
2150 like macro names defined in the configuration file.
2152 7. Added "dereference=" facility to LDAP.
2154 8. Two instances of the typo "uknown" in the source files are fixed.
2156 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2157 the Configure-Makefile script screwed up while processing it.
215910. Incorporated PCRE 4.4.
216111. The SMTP synchronization check was not operating right at the start of an
2162 SMTP session. For example, it could not catch a HELO sent before the client
2163 waited for the greeting. There is now a check for outstanding input at the
2164 point when the greeting is written. Because of the duplex, asynchronous
2165 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2166 way, but not yet received, when the check is performed.
216812. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2169 on TCP/IP sockets, because this apparently causes some broken clients to
2170 timeout.
217213. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2173 unchanged) from the Cygwin maintainer.
217514. The code for -bV that shows what is in the binary showed "mbx" when maildir
2176 was supported instead of testing for mbx. Effectively a typo.
217815. The spa authenticator server code was not checking that the input it
2179 received was valid base64.
218116. The debug output line for the "set" modifier in ACLs was not showing the
2182 name of the variable that was being set.
218417. Code tidy: the variable type "vtype_string" was never used. Removed it.
218618. Previously, a reference to $sender_host_name did not cause a DNS reverse
2187 lookup on its own. Something else was needed to trigger the lookup. For
2188 example, a match in host_lookup or the need for a host name in a host list.
2189 Now, if $sender_host_name is referenced and the host name has not yet been
2190 looked up, a lookup is performed. If the lookup fails, the variable remains
2191 empty, and $host_lookup_failed is set to "1".
219319. Added "eqi" as a case-independent comparison operator.
219520. The saslauthd authentication condition could segfault if neither service
2196 nor realm was specified.
219821. If an overflowing value such as "2048M" was set for message_size_limit, the
2199 error message that was logged was misleading, and incoming SMTP
2200 connections were dropped. The message is now more accurate, and temporary
2201 errors are given to SMTP connections.
220322. In some error situations (such as 21 above) Exim rejects all SMTP commands
2204 (except RSET) with a 421 error, until QUIT is received. However, it was
2205 failing to send a response to QUIT.
220723. The HELO ACL was being run before the code for helo_try_verify_hosts,
2208 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2209 ACL is now run after the helo_try_verify_hosts code.
221124. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2212 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2213 case-independent, so other case variants are also recognized). Apparently
2214 some systems use these upper case variants.
221625. If more than two messages were waiting for the same host, and a transport
2217 filter was specified for the transport, Exim sent two messages over the
2218 same TCP/IP connection, and then failed with "socket operation on non-
2219 socket" when it tried to send the third.
222126. Added Exim::debug_write and Exim::log_write for embedded Perl use.
222327. The extern definition of crypt16() in expand.c was not being excluded when
2224 the OS had its own crypt16() function.
222628. Added bounce_return_body as a new option, and bounce_return_size_limit
2227 as a preferred synonym for return_size_limit, both as an option and as an
2228 expansion variable.
223029. Added LIBS=-liconv to OS/Makefile-OSF1.
223230. Changed the default configuration ACL to relax the local part checking rule
2233 for addresses that are not in any local domains. For these addresses,
2234 slashes and pipe symbols are allowed within local parts, but the sequence
2235 /../ is explicitly forbidden.
223731. SPA server authentication was not clearing the challenge buffer before
2238 using it.
224032. log_message in a "warn" ACL statement was writing to the reject log as
2241 well as to the main log, which contradicts the documentation and doesn't
2242 seem right (because no rejection is happening). So I have stopped it.
224433. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2245 However, I am unable to do any testing of this.
224734. Fixed an infelicity in the appendfile transport. When checking directories
2248 for a mailbox, to see if any needed to be created, it was accidentally
2249 using path names with one or more superfluous leading slashes; tracing
2250 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
225235. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2253 contents are added to the log line that is written for every discarded
2254 recipient. (Previously a log_message setting was ignored.)
225636. The ${quote: operator now quotes the string if it is empty.
225837. The install script runs exim in order to find its version number. If for
2259 some reason other than non-existence or emptiness, which it checks, it
2260 could not run './exim', it was installing it with an empty version number,
2261 i.e. as "exim-". This error state is now caught, and the installation is
2262 aborted.
226438. An argument was missing from the function that creates an error message
2265 when Exim fails to connect to the socket for saslauthd authentication.
2266 This could cause Exim to crash, or give a corrupted message.
226839. Added isip, isip4, and isip6 to ${if conditions.
227040. The ACL variables $acl_xx are now saved with the message, and can be
2271 accessed later in routers, transports, and filters.
227341. The new lookup type nwildlsearch is like wildlsearch, except that the key
2274 strings in the file are not string-expanded.
227642. If a MAIL command specified a SIZE value that was too large to fit into an
2277 int variable, the check against message_size_limit failed. Such values are
2278 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2279 day this will have to be increased, but I don't think I want to be around
2280 when emails are that large.
2284Exim version 4.22
2287 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2288 iconv() is not standard in FreeBSD.
2290 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2291 IPv6 enabled. The observed symptom was a segmentation fault on return from
2292 the function os_common_find_running_interfaces() in src/os.c.
2294 3. In the check_special_case() function in daemon.c I had used "errno" as an
2295 argument name, which causes warnings on some systems. This was basically a
2296 typo, since it was named "eno" in the comments!
2298 4. The code that waits for the clock to tick (at a resolution of some fraction
2299 of a second) so as to ensure message-id uniqueness was always waiting for
2300 at least one whole tick, when it could have waited for less. [This is
2301 almost certainly not relevant at current processor speeds, where it is
2302 unlikely to ever wait at all. But we try to future-proof.]
2304 5. The function that sleeps for a time interval that includes fractions of a
2305 second contained a race. It did not block SIGALRM between setting the
2306 timer, and suspending (a couple of lines later). If the interval was short
2307 and the sigsuspend() was delayed until after it had expired, the suspension
2308 never ended. On busy systems this could lead to processes getting stuck for
2309 ever.
2311 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2312 process, before it forks any delivery processes. The open lookup caching
2313 mechanism meant that the open file or database connection was passed into
2314 the delivery process. The problem was that delivery processes always tidy
2315 up cached lookup data. This could cause a problem for the next delivery
2316 process started by the queue runner, because the external queue runner
2317 process does not know about the closure. So the next delivery process
2318 still has data in the lookup cache. In the case of a file lookup, there was
2319 no problem because closing a file descriptor in a subprocess doesn't affect
2320 the parent. However, if the lookup was caching a connection to a database,
2321 the connection was closed, and the second delivery process was likely to
2322 see errors such as "PGSQL: query failed: server closed the connection
2323 unexpectedly". The problem has been fixed by closing all cached lookups
2324 in a queue runner before running a delivery process.
2326 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2327 seem to have the "const" qualifier which it has on other OS. I've
2328 parameterised it.
2330 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2331 *of the same type* (client or server) with the same public name that an
2332 error should be diagnosed.
2334 9. When Exim looked up a host name for an IP address, but failed to find the
2335 original IP address when looking up the host name (a safety check), it
2336 output the message "<ip address> does not match any IP for NULL", which was
2337 confusing, to say the least. The bug was that the host name should have
2338 appeared instead of "NULL".
234010. Since release 3.03, if Exim is called by a uid other than root or the Exim
2341 user that is built into the binary, and the -C or -D options is used, root
2342 privilege is dropped before the configuration file is read. In addition,
2343 logging is switched to stderr instead of the normal log files. If the
2344 configuration then re-defines the Exim user, the unprivileged environment
2345 is probably not what is expected, so Exim logs a panic warning message (but
2346 proceeds).
2348 However, if deliver_drop_privilege is set, the unprivileged state may well
2349 be exactly what is intended, so the warning has been cut out in that case,
2350 and Exim is allowed to try to write to its normal log files.
2353Exim version 4.21
2356 1. smtp_return_error_details was not giving details for temporary sender
2357 or receiver verification errors.
2359 2. Diagnose a configuration error if two authenticators have the same public
2360 name.
2362 3. Exim used not to create the message log file for a message until the first
2363 delivery attempt. This could be confusing when incoming messages were held
2364 for policy or load reasons. The message log file is now created at the time
2365 the message is received, and an initial "Received" line is written to it.
2367 4. The automatically generated man page for command line options had a minor
2368 bug that caused no ill effects; however, a more serious problem was that
2369 the procedure for building the man page automatically didn't always
2370 operate. Consequently, release 4.20 contains an out-of-date version. This
2371 shouldn't happen again.
2373 5. When building Exim with embedded Perl support, the script that builds the
2374 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2375 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2377 6. The freeze_tell option was not being used for messages that were frozen on
2378 arrival, either by an ACL or by local_scan().
2380 7. Added the smtp_incomplete_transaction log selector.
2382 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2383 was accepting AUTH without a new EHLO.
2385 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2386 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2387 RFC.
238910. Logging of TCP/IP connections (when configured) now happens in the main
2390 daemon process instead of the child process, so that the TCP/IP connection
2391 count is more accurate (but it can never be perfect).
239311. The use of "drop" in a nested ACL was not being handled correctly in the
2394 outer ACL. Now, if condition failure induced by the nested "drop" causes
2395 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2396 or "require"), the connection is dropped.
239812. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2399 that yield "discard" can now be used with an "accept" or a "discard" verb,
2400 but an error is generated for any others (because I can't see a useful way
2401 to define what should happen).
240313. When an ACL is read dynamically from a file (or anywhere else), the lines
2404 are now processed in the same way as lines in the Exim configuration file.
2405 In particular, continuation lines are supported.
240714. Added the "dnslists = a.b.c!=n.n.n.n" feature.
240915. Added -ti meaning -t -i.
241116. Check for letters, digits, hyphens, and dots in the names of dnslist
2412 domains, and warn by logging if others are found.
241417. At least on BSD, alignment is not guarenteed for the array of ifreq's
2415 returned from GIFCONF when Exim is trying to find the list of interfaces on
2416 a host. The code in os.c has been modified to copy each ifreq to an aligned
2417 structure in all cases.
2419 Also, in some cases, the returned ifreq's were being copied to a 'struct
2420 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2421 means the last couple of bytes of an IPv6 address could be chopped if the
2422 ifreq contained only a normal sockaddr (14 bytes storage).
242418. Named domain lists were not supported in the hosts_treat_as_local option.
2425 An entry such as +xxxx was not recognized, and was treated as a literal
2426 domain name.
242819. Ensure that header lines added by a DATA ACL are included in the reject log
2429 if the ACL subsequently rejects the message.
243120. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2432 MD5 (which is deprecated).
243421. When testing a filter file using -bf, Exim was writing a message when it
2435 took the sender from a "From " line in the message, but it was not doing so
2436 when it took $return_path from a Return-Path: header line. It now does.
243822. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2439 with a valid header line field name (a series of printing characters
2440 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
244223. Changed "disc" in the source to "disk" to conform to the documentation and
2443 the book and for uniformity.
244524. Ignore Sendmail's -Ooption=value command line item.
244725. When execve() failed while trying to run a command in a pipe transport,
2448 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2449 could be confused with a return value of 69 from the command itself. This
2450 has been changed to 127, the value the shell returns if it is asked to run
2451 a non-existent command. The wording for the related log line suggests a
2452 non-existent command as the problem.
245426. If received_header_text expands to an empty string, do not add a Received:
2455 header line to the message. (Well, it adds a token one on the spool, but
2456 marks it "old" so that it doesn't get used or transmitted.)
245827. Installed eximstats 1.28 (addition of -nt option).
246028. There was no check for failure on the call to getsockname() in the daemon
2461 code. This can fail if there is a shortage of resources on the system, with
2462 ENOMEM, for example. A temporary error is now given on failure.
246429. Contrary to the C standard, it seems that in some environments, the
2465 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2466 program. Exim now does this explicitly; it affects the formatting of
2467 timestamps using strftime().
246930. If exiqsumm was given junk data, it threw up some uninitialized variable
2470 complaints. I've now initialized all the variables, to avoid this.
247232. Header lines added by a system filter were not being "seen" during
2473 transport-time rewrites.
247533. The info_callback() function passed to OpenSSL is set up with type void
2476 (*)(SSL *, int, int), as described somewhere. However, when calling the
2477 function (actually a macro) that sets it up, the type void(*)() is
2478 expected. I've put in a cast to prevent warnings from picky compilers.
248034. If a DNS black list lookup found a CNAME record, but there were no A
2481 records associated with the domain it pointed at, Exim crashed.
248335. If a DNS black list lookup returned more than one A record, Exim ignored
2484 all but the first. It now scans all returned addresses if a particular IP
2485 value is being sought. In this situation, the contents of the
2486 $dnslist_value variable are a list of all the addresses, separated by a
2487 comma and a space.
248936. Tightened up the rules for host name lookups using reverse DNS. Exim used
2490 to accept a host name and all its aliases if the forward lookup for any of
2491 them yielded the IP address of the incoming connection. Now it accepts only
2492 those names whose forward lookup yields the correct IP address. Any other
2493 names are discarded. This closes a loophole whereby a rogue DNS
2494 administrator could create reverse DNS records to break through a
2495 wildcarded host restriction in an ACL.
249737. If a user filter or a system filter that ran in a subprocess used any of
2498 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2499 the wrong values were passed to the pipe command ($thisaddress had the
2500 value of $0, $0 had the value of $1, etc). This bug was introduced by
2501 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
250338. Improved the line breaking for long SMTP error messages from ACLs.
2504 Previously, if there was no break point between 40 and 75 characters, Exim
2505 left the rest of the message alone. Two changes have been made: (a) I've
2506 reduced the minimum length to 35 characters; (b) if it can't find a break
2507 point between 35 and 75 characters, it looks ahead and uses the first one
2508 that it finds. This may give the occasional overlong line, but at least the
2509 remaining text gets split now.
251139. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2512 file descriptors might be low, and that setting 1000 would always raise it.
2513 It turns out that in some environments, the limit is already over 1000 and
2514 that lowering it causes trouble. So now Exim takes care not to decrease it.
251640. When delivering a message, the value of $return_path is set to $sender_
2517 address at the start of routing (routers may change the value). By an
2518 oversight, this default was not being set up when an address was tested by
2519 -bt or -bv, which affected the outcome if any router or filter referred to
2520 $return_path.
252241. The idea of the "warn" ACL verb is that it adds a header or writes to the
2523 log only when "message" or "log_message" are set. However, if one of the
2524 conditions was an address verification, or a call to a nested ACL, the
2525 messages generated by the underlying test were being passed through. This
2526 no longer happens. The underlying message is available in $acl_verify_
2527 message for both "message" and "log_message" expansions, so it can be
2528 passed through if needed.
253042. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2531 new expansion $bh_ to give the encoded byte string without charset
2532 translation. Translation happens only if iconv() is available; HAVE_ICONV
2533 indicates this at build time. HEADERS_CHARSET gives the charset to
2534 translate to; headers_charset can change it in the configuration, and
2535 "headers charset" can change it in an individual filter file.
253743. Now that we have a default RFC 2047 charset (see above), the code in Exim
2538 that creates RFC 2047 encoded "words" labels them as that charset instead
2539 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2540 expansion operator; (ii) when Exim creates a From: line for a local
2541 message; (iii) when a header line is rewritten to include a "phrase" part.
254344. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2544 buggy, causing it to skip the first lines of messages whose message ID
2545 ended in 'D'. This would not have bitten before Exim release 4.14, because
2546 message IDs were unlikely to end in 'D' before then. The effect was to have
2547 incorrect size information for certain domains.
254945. #include "config.h" was missing at the start of the crypt16.c module. This
2550 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2551 noticed.
255346. If there was a timeout during a "random" callout check, Exim treated it as
2554 a failure of the random address, and carried on sending RSET and the real
2555 address. If the delay was just some slowness somewhere, the response to the
2556 original RCPT would be taken as a response to RSET and so on, causing
2557 mayhem of various kinds.
255947. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2560 when I implemented it. It didn't allow for the fact that some option values
2561 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2562 the right test for positive values.
256448. Domain names in DNS records are case-independent. Exim always looks them up
2565 in lower case. Some resolvers return domain names in exactly the case they
2566 appear in the zone file, that is, they may contain uppercase letters. Not
2567 all resolvers do this - some return always lower case. Exim was treating a
2568 change of case by a resolver as a change of domain, similar to a widening
2569 of a domain abbreviation. This triggered its re-routing code and so it was
2570 trying to route what was effectively the same domain again. This normally
2571 caused routing to fail (because the router wouldn't handle the domain
2572 twice). Now Exim checks for this case specially, and just changes the
2573 casing of the domain that it ultimately uses when it transmits the message
2574 envelope.
257649. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2577 module.
257950. If a filter generated a file delivery with a non-absolute name (possible if
2580 no home directory exists for the router), the forbid_file option was not
2581 forbidding it.
258351. Added '&' feature to dnslists, to provide bit mask matching in addition to
2584 the existing equality matching.
258652. Exim was using ints instead of ino_t variables in some places where it was
2587 dealing with inode numbers.
258953. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2590 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2591 called TMPDIR, and if it finds it is different, it changes its value.
259354. The smtp_printf() function is now made available to local_scan() so
2594 additional output lines can be written before returning. There is also an
2595 smtp_fflush() function to enable the detection of a dropped connection.
2596 The variables smtp_input and smtp_batched_input are exported to
2597 local_scan().
259955. Changed the default runtime configuration: the message "Unknown user"
2600 has been removed from the ACL, and instead placed on the localuser router,
2601 using the cannot_route_message feature. This means that any verification
2602 failures that generate their own messages won't get overridden. Similarly,
2603 the "Unrouteable address" message that was in the ACL for unverifiable
2604 relay addresses has also been removed.
260656. Added hosts_avoid_esmtp to the smtp transport.
260857. The exicyclog script was not checking for the esoteric option
2609 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2610 will work only if exicyclog is run under the appropriate euid.
261258. Following a discussion on the list, the rules by which Exim recognises line
2613 endings on incoming messages have been changed. The -dropcr and drop_cr
2614 options are now no-ops, retained only for backwards compatibility. The
2615 following line terminators are recognized: LF CRLF CR. However, special
2616 processing applies to CR:
2618 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2619 nor a local message in the state where . is a terminator.
2621 (ii) If a bare CR is encountered in a header line, an extra space is added
2622 after the line terminator so as not to end the header. The reasoning
2623 behind this is that bare CRs in header lines are most likely either
2624 to be mistakes, or people trying to play silly games.
262659. The size of a message, as listed by "-bp" or in the Exim monitor window,
2627 was being incorrectly given as 18 bytes larger than it should have been.
2628 This is a VOB (very old bug).
263060. This may never have affected anything current, but just in case it has:
2631 When the local host is found other than at the start of a list of hosts,
2632 the local host, those with the same MX, and any that follow, are discarded.
2633 When the list in question was part of a longer list of hosts, the following
2634 hosts (not currently being processed) were also being discarded. This no
2635 longer happens. I'm not sure if this situation could ever has previously
2636 arisen.
263861. Added the "/MX" feature to lists of hosts in the manualroute and query
2639 program routers.
264162. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2642 header. This is something that is recommended in a new Internet Draft, and
2643 is something that is documented as being done by Sendmail. There are two
2644 possible values. For messages generated by the autoreply transport, Exim
2645 adds:
2647 Auto-Submitted: auto-replied
2649 whereas for all other generated messages (e.g. bounces) it adds
2651 Auto-Submitted: auto-generated
265363. The "personal" condition in filters now includes a test for the
2654 Auto-Submitted: header. If it contains the string "auto-" the message it
2655 not considered personal.
265764. Added rcpt_include_affixes as a generic transport option.
265965. Added queue_only_override (default true).
266166. Added the syslog_duplication option.
266367. If what should have been the first header line of a message consisted of
2664 a space followed by a colon, Exim was mis-interpreting it as a header line.
2665 It isn't of course - it is syntactically invalid and should therefore be
2666 treated as the start of the message body. The misbehaviour could have
2667 caused a number of strange effects, including loss of data in subsequent
2668 header lines, and spool format errors.
267068. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2671 client host had authenticated. This control can now be exercised by an ACL
2672 for more flexibility.
267469. By default, callouts do not happen when testing with -bh. There is now a
2675 variant, -bhc, which does actually run the callout code, including
2676 consulting and updating the callout cache.
267870. Added support for saslauthd authentication, courtesy of Alexander
2679 Sabourenkov.
268171. If statvfs() failed on the spool or log directories while checking their
2682 size for availability, Exim confusingly gave the error "space shortage".
2683 Furthermore, in debugging mode it crashed with a floating point exception.
2684 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2685 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2686 serious problem, Exim now writes to the main and panic logs when this
2687 happens, with details of the failure. It then refuses to accept the
2688 incoming message, giving the message "spool directory problem" or "log
2689 directory problem" with a 421 code for SMTP messages.
269172. When Exim is about to re-exec itself, it ensures that the file descriptors
2692 0, 1, and 2 exist, because some OS complain for execs without them (see
2693 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2694 descriptors. However, the code omitted to check that the open succeeded,
2695 causing mysterious errors if for some reason the permissions on /dev/null
2696 got screwed. Now Exim writes a message to the main and panic logs, and
2697 bombs out if it can't open /dev/null.
269973. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2700 interact so that it is all more flexible. It is supposed to remain
2701 backwards compatible. Also added extra_local_interfaces.
270374. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2704 to bomb out with an assertion failure - to the client this appears as a
2705 connection drop. This problem occurs in the part of the code that was taken
2706 from the Samba project. Fortunately, the assertion is in a very simple
2707 function, so I have fixed this by reproducing the function inline in the
2708 one place where it is called, and arranging for authentication to fail
2709 instead of killing the process with assert().
271175. The SPA client code was not working when the server requested OEM rather
2712 than Unicode encoding.
271476. Added code to make require_files with a specific uid setting more usable in
2715 the case where statting the file as root fails - usually a non-root-mounted
2716 NFS file system. When this happens and the failure is EACCES, Exim now
2717 forks a subprocess and does the per-uid checking as the relevant uid.
271977. Added process_log_path.
272178. If log_file_path was not explicitly set, a setting of check_log_space or
2722 check_log_inodes was ignored.
272479. If a space check for the spool or log partitions fails, the incident is now
2725 logged. Of course, in the latter case the data may get lost...
272780. Added the %p formatting code to string_format() so that it can be used to
2728 print addresses in debug_print(). Adjusted all the address printing in the
2729 debugging in store.c to use %p rather than %d.
273181. There was a concern that a line of code in smtp_in.c could overflow a
2732 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2733 initially expressed, the concern was not well-founded, because trailing
2734 spaces are removed early. However, if the trailing spaces were followed by
2735 a NULL, they did not get removed, so the overflow was possible. Two fixes
2736 were applied:
2738 (a) I re-wrote the offending code in a cleaner fashion.
2739 (b) If an incoming SMTP command contains a NULL character, it is rejected
2740 as invalid.
274282. When Exim changes uid/gid to the Exim user at daemon start time, it now
2743 runs initgroups(), so that if the Exim user is in any additional groups,
2744 they will be used during message reception.
2747Exim version 4.20
2750The change log for 4.20 and earlier releases has been archived.