1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.92 2005/03/15 15:36:41 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 TK/01. Added Yahoo DomainKeys support via libdomainkeys. See
11 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
13 TK/02. Fix ACL "control" statment not being available in MIME ACL.
15 TK/03. Fix ACL "regex" condition not being available in MIME ACL.
17 PH/01. Installed a patch from the Sieve maintainer that allows -bf to be used
18 to test Sieve filters that use "vacation".
20 PH/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
28 PH/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
34 PH/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.
43 PH/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.
46 PH/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.)
54 A note about Exim versions 4.44 and 4.50
55 ----------------------------------------
57 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
58 changes of various kinds. As a consequence, a big documentation update was
59 needed. This delayed the release for rather longer than seemed good, especially
60 in the light of a couple of (minor) security issues. Therefore, the changes
61 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
62 release. So 4.44 and 4.50 are in effect two different branches that both start
65 I have left the 4.50 change log unchanged; it contains all the changes since
66 4.43. The change log for 4.44 is below; many of its items are identical to
67 those for 4.50. This seems to be the most sensible way to preserve the
68 historical information.
74 1. Minor wording change to the doc/README.SIEVE file.
76 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
77 computation of the current number of files was incorrect.
79 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
80 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
83 4. Give more explanation in the error message when the command for a transport
84 filter fails to execute.
86 5. There are several places where Exim runs a non-Exim command in a
87 subprocess. The SIGUSR1 signal should be disabled for these processes. This
88 was being done only for the command run by the queryprogram router. It is
89 now done for all such subprocesses. The other cases are: ${run, transport
90 filters, and the commands run by the lmtp and pipe transports.
92 6. Added CONFIGURE_GROUP build-time option.
94 7. Some older OS have a limit of 256 on the maximum number of file
95 descriptors. Exim was using setrlimit() to set 1000 as a large value
96 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
97 systems. I've change it so that if it can't get 1000, it tries for 256.
99 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
100 was an oversight, and furthermore, ever since the addition of extra
101 controls (e.g. 4.43/32), the checks on when to allow different forms of
102 "control" were broken. There should now be diagnostics for all cases when a
103 control that does not make sense is encountered.
105 9. Added the /retain_sender option to "control=submission".
107 10. $recipients is now available in the predata ACL (oversight).
109 11. Tidy the search cache before the fork to do a delivery from a message
110 received from the command line. Otherwise the child will trigger a lookup
111 failure and thereby defer the delivery if it tries to use (for example) a
112 cached ldap connection that the parent has called unbind on.
114 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
115 of $address_data from the recipient verification was clobbered by the
118 13. The value of address_data from a sender verification is now available in
119 $sender_address_data in subsequent conditions in the ACL statement.
121 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
123 15. Added a new option "connect=<time>" to callout options, to set a different
126 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
127 was its contents. (It was OK if the option was not defined at all.)
129 17. A "Completed" log line is now written for messages that are removed from
130 the spool by the -Mrm option.
132 18. New variables $sender_verify_failure and $recipient_verify_failure contain
133 information about exactly what failed.
135 19. Added -dd to debug only the daemon process.
137 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
138 handles timeouts, both on the server side and network timeouts. Renamed the
139 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
141 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
142 It is now set to "smtps".
144 22. $host_address is now set to the target address during the checking of
147 23. When checking ignore_target_hosts for an ipliteral router, no host name was
148 being passed; this would have caused $sender_host_name to have been used if
149 matching the list had actually called for a host name (not very likely,
150 since this list is usually IP addresses). A host name is now passed as
153 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
154 code that specifies a non-restarting handler (typically sigaction() in
155 modern systems) in an attempt to fix a rare and obscure crash bug.
157 25. Narrowed the window for a race in the daemon that could cause it to ignore
158 SIGCHLD signals. This is not a major problem, because they are used only to
159 wake it up if nothing else does.
161 26. A malformed maildirsize file could cause Exim to calculate negative values
162 for the mailbox size or file count. Odd effects could occur as a result.
163 The maildirsize information is now recalculated if the size or filecount
166 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
167 support for a long time. Removed HAVE_SYS_VFS_H.
169 28. Installed the latest version of exipick from John Jetmore.
171 29. In an address list, if the pattern was not a regular expression, an empty
172 subject address (from a bounce message) matched only if the pattern was an
173 empty string. Non-empty patterns were not even tested. This was the wrong
174 because it is perfectly reasonable to use an empty address as part of a
175 database query. An empty address is now tested by patterns that are
176 lookups. However, all the other forms of pattern expect the subject to
177 contain a local part and a domain, and therefore, for them, an empty
178 address still always fails if the pattern is not itself empty.
180 30. Exim went into a mad DNS loop when attempting to do a callout where the
181 host was specified on an smtp transport, and looking it up yielded more
184 31. Re-factored the code for checking spool and log partition space into a
185 function that finds that data and another that does the check. The former
186 is then used to implement four new variables: $spool_space, $log_space,
187 $spool_inodes, and $log_inodes.
189 32. The RFC2047 encoding function was originally intended for short strings
190 such as real names; it was not keeping to the 75-character limit for
191 encoded words that the RFC imposes. It now respects the limit, and
192 generates multiple encoded words if necessary. To be on the safe side, I
193 have increased the buffer size for the ${rfc2047: expansion operator from
196 33. It is now permitted to omit both strings after an "if" condition; if the
197 condition is true, the result is "true". As before, when the second string
198 is omitted, a false condition yields an empty string. This makes it less
199 cumbersome to write custom ACL and router conditions.
201 34. Failure to deliver a bounce message always caused it to be frozen, even if
202 there was an errors_to setting on the router. The errors_to setting is now
205 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
206 canonical form (fully expanded) before being placed in
207 $sender_host_address.
209 36. The table in the code that translates DNS record types into text (T_A to
210 "A" for instance) was missing entries for NS and CNAME. It is just possible
211 that this could have caused confusion if both these types were looked up
212 for the same domain, because the text type is used as part of Exim's
213 per-process caching. But the chance of anyone hitting this buglet seems
216 37. The dnsdb lookup has been extended in a number of ways.
218 (1) There is a new type, "zns", which walks up the domain tree until it
219 finds some nameserver records. It should be used with care.
221 (2) There is a new type, "mxh", which is like "mx" except that it returns
222 just the host names, not the priorities.
224 (3) It is now possible to give a list of domains (or IP addresses) to be
225 looked up. The behaviour when one of the lookups defers can be
226 controlled by a keyword.
228 (4) It is now possible to specify the separator character for use when
229 multiple records are returned.
231 38. The dnslists ACL condition has been extended: it is now possible to supply
232 a list of IP addresses and/or domains to be looked up in a particular DNS
235 39. Added log_selector=+queue_time_overall.
237 40. When running the queue in the test harness, wait just a tad after forking a
238 delivery process, to get repeatability of debugging output.
240 41. Include certificate and key file names in error message when GnuTLS fails
241 to set them up, because the GnuTLS error message doesn't include the name
242 of the failing file when there is a problem reading it.
244 42. Allow both -bf and -bF in the same test run.
246 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
248 44. The "Exiscan patch" is now merged into the mainline Exim source.
250 45. Sometimes the final signoff response after QUIT could fail to get
251 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
252 < 0 before doing a fflush(). This bug looks as though it goes back to the
253 introduction of TLS in release 3.20, but "sometimes" must have been rare
254 because the tests only now provoked it.
256 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
257 (this can affect the format of dates).
259 47. exim_tidydb, when checking for the continued existence of a message for
260 which it has found a message-specific retry record, was not finding
261 messages that were in split spool directories. Consequently, it was
262 deleting retry records that should have stayed in existence.
264 48. Steve fixed some bugs in eximstats.
266 49. The SPA authentication driver was not abandoning authentication and moving
267 on to the next authenticator when an expansion was forced to fail,
268 contradicting the general specification for all authenticators. Instead it
269 was generating a temporary error. It now behaves as specified.
271 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
272 (the order specifies the preference for clients). The order is now AES256,
273 AES128, 3DES, ARCFOUR128.
275 51. Small patch to Sieve code - explicitly set From: when generating an
278 52. Exim crashed if a remote delivery caused a very long error message to be
279 recorded - for instance if somebody sent an entire SpamAssassin report back
280 as a large number of 550 error lines. This bug was coincidentally fixed by
281 increasing the size of one of Exim's internal buffers (big_buffer) that
282 happened as part of the Exiscan merge. However, to be on the safe side, I
283 have made the code more robust (and fixed the comments that describe what
286 53. Now that there can be additional text after "Completed" in log lines (if
287 the queue_time_overall log selector is set), a one-byte patch to exigrep
288 was needed to allow it to recognize "Completed" as not the last thing in
291 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
292 patch that reportedly fixes this has been added. I am not expert enough to
293 create a test for it. This is what the patch creator wrote:
295 "I found a little strange behaviour of ldap code when working with
296 Windows 2003 AD Domain, where users was placed in more than one
297 Organization Units. When I tried to give exim partial DN, the exit code
298 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
299 But simultaneously result of request was absolutely normal ldap result,
300 so I produce this patch..."
302 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
303 REFERENCE, so I have modified the code to exclude the patch when that macro
306 55. Some experimental protocols are using DNS PTR records for new purposes. The
307 keys for these records are domain names, not reversed IP addresses. The
308 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
309 leaves it alone. Component reversal etc. now happens only for IP addresses.
312 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
314 57. Double the size of the debug message buffer (to 2048) so that more of very
315 long debug lines gets shown.
317 58. The exicyclog utility now does better if the number of log files to keep
318 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
320 59. Two changes related to the smtp_active_hostname option:
322 (1) $smtp_active_hostname is now available as a variable.
323 (2) The default for smtp_banner uses $smtp_active_hostname instead
324 of $primary_hostname.
326 60. The host_aton() function is supposed to be passed a string that is known
327 to be a valid IP address. However, in the case of IPv6 addresses, it was
328 not checking this. This is a hostage to fortune. Exim now panics and dies
329 if the condition is not met. A case was found where this could be provoked
330 from a dnsdb PTR lookup with an IPv6 address that had more than 8
331 components; fortuitously, this particular loophole had already been fixed
332 by change 4.50/55 above.
334 If there are any other similar loopholes, the new check in host_aton()
335 itself should stop them being exploited. The report I received stated that
336 data on the command line could provoke the exploit when Exim was running as
337 exim, but did not say which command line option was involved. All I could
338 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
342 61. There was a buffer overflow vulnerability in the SPA authentication code
343 (which came originally from the Samba project). I have added a test to the
344 spa_base64_to_bits() function which I hope fixes it.
347 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
348 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
350 63. The daemon start-up calls getloadavg() while still root for those OS that
351 need the first call to be done as root, but it missed one case: when
352 deliver_queue_load_max is set with deliver_drop_privilege. This is
353 necessary for the benefit of the queue runner, because there is no re-exec
354 when deliver_drop_privilege is set.
356 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
359 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
360 was in use, was not putting the data itself into the right store pool;
361 consequently, it could be overwritten for a subsequent message in the same
362 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
365 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
367 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
368 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
369 the function were treating the return as a boolean value, which happened to
370 work because 0=false and not-0=true, but is not correct code.
372 68. The host_aton() function was not handling scoped IPv6 addresses (those
373 with, for example, "%eth0" on the end) correctly.
375 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
376 negated items (that is, ~something) in unsigned ints. Some compilers
377 apparently mutter when there is no cast.
379 70. If an address verification called from an ACL failed, and did not produce a
380 user-specific message (i.e. there was only a "system" message), nothing was
381 put in $acl_verify_message. In this situation, it now puts the system
384 71. Change 4.23/11 added synchronization checking at the start of an SMTP
385 session; change 4.31/43 added the unwanted input to the log line - except
386 that it did not do this in the start of session case. It now does.
388 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
389 This is wrong and can cause the other end to generate a synchronization
390 error if it is another Exim or anything else that does the synchronization
391 check. A QUIT command is no longer sent after a timeout.
393 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
396 74. The defer_ok option of callout verification was not working if it was used
397 when verifying addresses in header lines, that is, for this case:
399 verify = header_sender/callout=defer_ok
401 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
402 those file descriptors could be used for SMTP connections. If anything
403 wrote to stderr (the example that came up was "warn" in embedded Perl), it
404 could be sent to the SMTP client, causing chaos. The daemon now opens
405 stdin, stdout, and stderr to /dev/null when it puts itself into the
408 76. Arrange for output from Perl's "warn" command to be written to Exim's main
409 log by default. The user can override this with suitable Perl magic.
411 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
412 the log message when discard triggers, was not working for the DATA ACL or
413 for the non-SMTP ACL.
415 78. Error message wording change in sieve.c.
417 79. If smtp_accept_max_per_host was set, the number of connections could be
418 restricted to fewer than expected, because the daemon was trying to set up
419 a new connection before checking whether the processes handling previous
420 connections had finished. The check for completed processes is now done
421 earlier. On busy systems, this bug wouldn't be noticed because something
422 else would have woken the daemon, and it would have reaped the completed
425 80. If a message was submitted locally by a user whose login name contained one
426 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
427 It caused a spool format error. I have fixed the spool reading code. A
428 related problem was that the "from" clause in the Received: line became
429 illegal because of the space(s). It is now covered by ${quote_local_part.
431 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
434 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
436 83. Patch from Sieve maintainer to fix unterminated string problem in
439 84. Some minor changes to the Linux configuration files to help with other
440 OS variants using glibc.
442 85. One more patch for Sieve to update vacation handling to latest spec.
445 ----------------------------------------------------
446 See the note above about the 4.44 and 4.50 releases.
447 ----------------------------------------------------
453 1. Change 4.43/35 introduced a bug that caused file counts to be
454 incorrectly computed when quota_filecount was set in an appendfile
457 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
458 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
461 3. Give more explanation in the error message when the command for a transport
462 filter fails to execute.
464 4. There are several places where Exim runs a non-Exim command in a
465 subprocess. The SIGUSR1 signal should be disabled for these processes. This
466 was being done only for the command run by the queryprogram router. It is
467 now done for all such subprocesses. The other cases are: ${run, transport
468 filters, and the commands run by the lmtp and pipe transports.
470 5. Some older OS have a limit of 256 on the maximum number of file
471 descriptors. Exim was using setrlimit() to set 1000 as a large value
472 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
473 systems. I've change it so that if it can't get 1000, it tries for 256.
475 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
476 was an oversight, and furthermore, ever since the addition of extra
477 controls (e.g. 4.43/32), the checks on when to allow different forms of
478 "control" were broken. There should now be diagnostics for all cases when a
479 control that does not make sense is encountered.
481 7. $recipients is now available in the predata ACL (oversight).
483 8. Tidy the search cache before the fork to do a delivery from a message
484 received from the command line. Otherwise the child will trigger a lookup
485 failure and thereby defer the delivery if it tries to use (for example) a
486 cached ldap connection that the parent has called unbind on.
488 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
489 of $address_data from the recipient verification was clobbered by the
492 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
493 was its contents. (It was OK if the option was not defined at all.)
495 11. A "Completed" log line is now written for messages that are removed from
496 the spool by the -Mrm option.
498 12. $host_address is now set to the target address during the checking of
501 13. When checking ignore_target_hosts for an ipliteral router, no host name was
502 being passed; this would have caused $sender_host_name to have been used if
503 matching the list had actually called for a host name (not very likely,
504 since this list is usually IP addresses). A host name is now passed as
507 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
508 code that specifies a non-restarting handler (typically sigaction() in
509 modern systems) in an attempt to fix a rare and obscure crash bug.
511 15. Narrowed the window for a race in the daemon that could cause it to ignore
512 SIGCHLD signals. This is not a major problem, because they are used only to
513 wake it up if nothing else does.
515 16. A malformed maildirsize file could cause Exim to calculate negative values
516 for the mailbox size or file count. Odd effects could occur as a result.
517 The maildirsize information is now recalculated if the size or filecount
520 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
521 support for a long time. Removed HAVE_SYS_VFS_H.
523 18. Updated exipick to current release from John Jetmore.
525 19. Allow an empty sender to be matched against a lookup in an address list.
526 Previously the only cases considered were a regular expression, or an
529 20. Exim went into a mad DNS lookup loop when doing a callout where the
530 host was specified on the transport, if the DNS lookup yielded more than
533 21. The RFC2047 encoding function was originally intended for short strings
534 such as real names; it was not keeping to the 75-character limit for
535 encoded words that the RFC imposes. It now respects the limit, and
536 generates multiple encoded words if necessary. To be on the safe side, I
537 have increased the buffer size for the ${rfc2047: expansion operator from
540 22. Failure to deliver a bounce message always caused it to be frozen, even if
541 there was an errors_to setting on the router. The errors_to setting is now
544 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
545 canonical form (fully expanded) before being placed in
546 $sender_host_address.
548 24. Updated eximstats to version 1.33
550 25. Include certificate and key file names in error message when GnuTLS fails
551 to set them up, because the GnuTLS error message doesn't include the name
552 of the failing file when there is a problem reading it.
554 26. Expand error message when OpenSSL has problems setting up cert/key files.
557 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
558 (this can affect the format of dates).
560 28. exim_tidydb, when checking for the continued existence of a message for
561 which it has found a message-specific retry record, was not finding
562 messages that were in split spool directories. Consequently, it was
563 deleting retry records that should have stayed in existence.
565 29. eximstats updated to version 1.35
566 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
567 1.35 - bugfix such that pie charts by volume are generated correctly
569 30. The SPA authentication driver was not abandoning authentication and moving
570 on to the next authenticator when an expansion was forced to fail,
571 contradicting the general specification for all authenticators. Instead it
572 was generating a temporary error. It now behaves as specified.
574 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
575 (the order specifies the preference for clients). The order is now AES256,
576 AES128, 3DES, ARCFOUR128.
578 31. Small patch to Sieve code - explicitly set From: when generating an
581 32. Exim crashed if a remote delivery caused a very long error message to be
582 recorded - for instance if somebody sent an entire SpamAssassin report back
583 as a large number of 550 error lines. This bug was coincidentally fixed by
584 increasing the size of one of Exim's internal buffers (big_buffer) that
585 happened as part of the Exiscan merge. However, to be on the safe side, I
586 have made the code more robust (and fixed the comments that describe what
589 33. Some experimental protocols are using DNS PTR records for new purposes. The
590 keys for these records are domain names, not reversed IP addresses. The
591 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
592 leaves it alone. Component reversal etc. now happens only for IP addresses.
595 34. The host_aton() function is supposed to be passed a string that is known
596 to be a valid IP address. However, in the case of IPv6 addresses, it was
597 not checking this. This is a hostage to fortune. Exim now panics and dies
598 if the condition is not met. A case was found where this could be provoked
599 from a dnsdb PTR lookup with an IPv6 address that had more than 8
600 components; fortuitously, this particular loophole had already been fixed
601 by change 4.50/55 or 4.44/33 above.
603 If there are any other similar loopholes, the new check in host_aton()
604 itself should stop them being exploited. The report I received stated that
605 data on the command line could provoke the exploit when Exim was running as
606 exim, but did not say which command line option was involved. All I could
607 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
611 35. There was a buffer overflow vulnerability in the SPA authentication code
612 (which came originally from the Samba project). I have added a test to the
613 spa_base64_to_bits() function which I hope fixes it.
616 36. The daemon start-up calls getloadavg() while still root for those OS that
617 need the first call to be done as root, but it missed one case: when
618 deliver_queue_load_max is set with deliver_drop_privilege. This is
619 necessary for the benefit of the queue runner, because there is no re-exec
620 when deliver_drop_privilege is set.
622 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
623 was in use, was not putting the data itself into the right store pool;
624 consequently, it could be overwritten for a subsequent message in the same
625 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
628 38. Sometimes the final signoff response after QUIT could fail to get
629 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
630 < 0 before doing a fflush(). This bug looks as though it goes back to the
631 introduction of TLS in release 3.20, but "sometimes" must have been rare
632 because the tests only now provoked it.
638 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
639 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
640 Now it returns an integer. A number of calls were still expecting a T/F
641 return. Fortuitously, in all cases, the tests worked in OK situations,
642 which is the norm. However, things would have gone wrong on any write
643 failures on the smtp file descriptor. This function is used when sending
644 messages over SMTP and also when doing verify callouts.
646 2. When Exim is called to do synchronous delivery of a locally submitted
647 message (the -odf or -odi options), it no longer closes stderr before doing
650 3. Implemented the mua_wrapper option.
652 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
654 5. Implemented the functions header_remove(), header_testname(),
655 header_add_at_position(), and receive_remove_recipient(), and exported them
658 6. If an ACL "warn" statement specified the addition of headers, Exim already
659 inserted X-ACL-Warn: at the start if there was no header name. However, it
660 was not making this test for the second and subsequent header lines if
661 there were newlines in the string. This meant that an invalid header could
662 be inserted if Exim was badly configured.
664 7. Allow an ACL "warn" statement to add header lines at the start or after all
665 the Received: headers, as well as at the end.
667 8. Added the rcpt_4xx retry error code.
669 9. Added postmaster_mailfrom=xxx to callout verification option.
671 10. Added mailfrom=xxxx to the callout verification option, for verify=
674 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
675 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
677 12. Inserted some casts to stop certain compilers warning when using pointer
678 differences as field lengths or precisions in printf-type calls (mostly
679 affecting debugging statements).
681 13. Added optional readline() support for -be (dynamically loaded).
683 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
684 same clock tick as a message's arrival, so that its received time was the
685 same as the "first fail" time on the retry record, and that message
686 remained on the queue past the ultimate address timeout, every queue runner
687 would try a delivery (because it was past the ultimate address timeout) but
688 after another failure, the ultimate address timeout, which should have then
689 bounced the address, did not kick in. This was a "< instead of <=" error;
690 in most cases the first failure would have been in the next clock tick
691 after the received time, and all would be well.
693 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
694 being recognized when the domain list was tested by the match_domain
695 condition in an expansion string.
697 16. Added the ${str2b64: operator.
699 17. Exim was always calling setrlimit() to set a large limit for the number of
700 processes, without checking whether the existing limit was already
701 adequate. (It did check for the limit on file descriptors.) Furthermore,
702 errors from getrlimit() and setrlimit() were being ignored. Now they are
703 logged to the main and panic logs, but Exim does carry on, to try to do its
704 job under whatever limits there are.
706 18. Imported PCRE 5.0.
708 19. Trivial typo in log message " temporarily refused connection" (the leading
711 20. If the log selector return_path_on_delivery was set and an address was
712 redirected to /dev/null, the delivery process crashed because it assumed
713 that a return path would always be set for a "successful" delivery. In this
714 case, the whole delivery is bypassed as an optimization, and therefore no
717 21. Internal re-arrangement: the function for sending a challenge and reading
718 a response while authentication was assuming a zero-terminated challenge
719 string. It's now changed to take a pointer and a length, to allow for
720 binary data in such strings.
722 22. Added the cyrus_sasl authenticator (code supplied by MBM).
724 23. Exim was not respecting finduser_retries when seeking the login of the
725 uid under which it was called; it was always trying 10 times. (The default
726 setting of finduser_retries is zero.) Also, it was sleeping after the final
727 failure, which is pointless.
729 24. Implemented tls_on_connect_ports.
731 25. Implemented acl_smtp_predata.
733 26. If the domain in control=submission is set empty, Exim assumes that the
734 authenticated id is a complete email address when it generates From: or
735 Sender: header lines.
737 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
738 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
739 chown and chgrp in /bin and hostname in /usr/bin.
741 28. Exim was keeping the "process log" file open after each use, just as it
742 does for the main log. This opens the possibility of it remaining open for
743 long periods when the USR1 signal hits a daemon. Occasional processlog
744 errors were reported, that could have been caused by this. Anyway, it seems
745 much more sensible not to leave this file open at all, so that is what now
748 29. The long-running daemon process does not normally write to the log once it
749 has entered its main loop, and it closes the log before doing so. This is
750 so that log files can straightforwardly be renamed and moved. However,
751 there are a couple of unusual error situations where the daemon does write
752 log entries, and I had neglected to close the log afterwards.
754 30. The text of an SMTP error response that was received during a remote
755 delivery was being truncated at 512 bytes. This is too short for some of
756 the long messages that one sometimes sees. I've increased the limit to
759 31. It is now possible to make retry rules that apply only when a message has a
760 specific sender, in particular, an empty sender.
762 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
763 it possible to be selective about when SMTP synchronization is enforced.
765 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
767 32. Implemented hosts_connection_nolog.
769 33. Added an ACL for QUIT.
771 34. Setting "delay_warning=" to disable warnings was not working; it gave a
774 35. Added mailbox_size and mailbox_filecount to appendfile.
776 36. Added control = no_multiline_responses to ACLs.
778 37. There was a bug in the logic of the code that waits for the clock to tick
779 in the case where the clock went backwards by a substantial amount such
780 that the microsecond fraction of "now" was more than the microsecond
781 fraction of "then" (but the whole seconds number was less).
783 38. Added support for the libradius Radius client library this is found on
784 FreeBSD (previously only the radiusclient library was supported).
790 1. When certain lookups returned multiple values in the form name=value, the
791 quoting of the values was not always being done properly. Specifically:
792 (a) If the value started with a double quote, but contained no whitespace,
794 (b) If the value contained whitespace other than a space character (i.e.
795 tabs or newlines or carriage returns) it was not quoted.
796 This fix has been applied to the mysql and pgsql lookups by writing a
797 separate quoting function and calling it from the lookup code. The fix
798 should probably also be applied to nisplus, ibase and oracle lookups, but
799 since I cannot test any of those, I have not disturbed their existing code.
801 2. A hit in the callout cache for a specific address caused a log line with no
802 reason for rejecting RCPT. Now it says "Previous (cached) callout
803 verification failure".
805 3. There was an off-by-one bug in the queryprogram router. An over-long
806 return line was truncated at 256 instead of 255 characters, thereby
807 overflowing its buffer with the terminating zero. As well as fixing this, I
808 have increased the buffer size to 1024 (and made a note to document this).
810 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
811 when Exim is waiting for an SMTP response from a remote server, Exim
812 restarts its select() call on the socket, thereby resetting its timeout.
813 This is not a problem when such interrupts are rare. Somebody set up a cron
814 job to run exiwhat every 2 minutes, which is less than the normal select()
815 timeout (5 or 10 minutes). This meant that the select() timeout never
816 kicked in because it was always reset. I have fixed this by comparing the
817 time when an interrupt arrives with the time at the start of the first call
818 to select(). If more time than the timeout has elapsed, the interrupt is
819 treated as a timeout.
821 5. Some internal re-factoring in preparation for the addition of Sieve
822 extensions (by MH). In particular, the "personal" test is moved to a
823 separate function, and given an option for scanning Cc: and Bcc: (which is
824 not set for Exim filters).
826 6. When Exim created an email address using the login of the caller as the
827 local part (e.g. when creating a From: or Sender: header line), it was not
828 quoting the local part when it contained special characters such as @.
830 7. Installed new OpenBSD configuration files.
832 8. Reworded some messages for syntax errors in "and" and "or" conditions to
833 try to make them clearer.
835 9. Callout options, other than the timeout value, were being ignored when
836 verifying sender addresses in header lines. For example, when using
838 verify = header_sender/callout=no_cache
840 the cache was (incorrectly) being used.
842 10. Added a missing instance of ${EXE} to the exim_install script; this affects
843 only the Cygwin environment.
845 11. When return_path_on_delivery was set as a log selector, if different remote
846 addresses in the same message used different return paths and parallel
847 remote delivery occurred, the wrong values would sometimes be logged.
848 (Whenever a remote delivery process finished, the return path value from
849 the most recently started remote delivery process was logged.)
851 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
852 lines when AUTH and/or TLS are in use. This is the "received protocol"
853 field. Exim used to use "asmtp" for authenticated SMTP, without any
854 indication (in the protocol name) for TLS use. Now it follows the RFC and
855 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
856 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
857 names appear in log lines as well as in Received: header lines.
859 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
860 extensions, and comparison tests, and to fix some bugs.
862 14. Changes to the "personal" filter test:
864 (1) The test was buggy in that it was just doing the equivalent of
865 "contains" tests on header lines. For example, if a user's address was
866 anne@some.where, the "personal" test would incorrectly be true for
868 To: susanne@some.where
870 This test is now done by extracting each address from the header in turn,
871 and checking the entire address. Other tests that are part of "personal"
872 are now done using regular expressions (for example, to check local parts
873 of addresses in From: header lines).
875 (2) The list of non-personal local parts in From: addresses has been
876 extended to include "listserv", "majordomo", "*-request", and "owner-*",
877 taken from the Sieve specification recommendations.
879 (3) If the message contains any header line starting with "List-" it is
880 treated as non-personal.
882 (4) The test for "circular" in the Subject: header line has been removed
883 because it now seems ill-conceived.
885 15. Minor typos in src/EDITME comments corrected.
887 16. Installed latest exipick from John Jetmore.
889 17. If headers_add on a router specified a text string that was too long for
890 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
891 of string_sprintf() is now avoided.
893 18. $message_body_size was not set (it was always zero) when running the DATA
894 ACL and the local_scan() function.
896 19. For the "mail" command in an Exim filter, no default was being set for
897 the once_repeat time, causing a random time value to be used if "once" was
898 specified. (If the value happened to be <= 0, no repeat happened.) The
899 default is now 0s, meaning "never repeat". The "vacation" command was OK
900 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
901 (I found it when inspecting the code).
903 20. There is now an overall timeout for performing a callout verification. It
904 defaults to 4 times the callout timeout, which applies to individual SMTP
905 commands during the callout. The overall timeout applies when there is more
906 than one host that can be tried. The timeout is checked before trying the
907 next host. This prevents very long delays if there are a large number of
908 hosts and all are timing out (e.g. when the network connections are timing
909 out). The value of the overall timeout can be changed by specifying an
910 additional sub-option for "callout", called "maxwait". For example:
912 verify = sender/callout=5s,maxwait=20s
914 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
915 to the end before writing, but this should make it even safer).
917 22. Exim was forgetting that it had advertised PIPELINING for the second and
918 subsequent messages on an SMTP connection. It was also not resetting its
919 memory on STARTTLS and an internal HELO.
921 23. When Exim logs an SMTP synchronization error within a session, it now
922 records whether PIPELINING has been advertised or not.
924 24. Added 3 instances of "(long int)" casts to time_t variables that were being
925 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
926 rather than long int.
928 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
930 26. Added the never_mail option to autoreply.
936 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
937 crash if the getsockname() call failed; this can happen if a connection is
938 closed very soon after it is established. The problem was simply in the
939 order in which certain operations were done, causing Exim to try to write
940 to the SMTP stream before it had set up the file descriptor. The bug has
941 been fixed by making things happen in the correct order.
947 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
948 before the connection was closed, thus losing the rejection response.
950 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
951 some early Solaris releases, but causes trouble in current releases where
952 socklen_t is defined.
954 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
957 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
960 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
961 that releasing the top of it at the end releases what was used for sub-
962 expansions (unless the block got too big). However, discard this block if
963 the first thing is a variable or header, so that we can use its block when
964 it is dynamic (useful for very large $message_headers, for example).
966 6. Lookups now cache *every* query, not just the most recent. A new, separate
967 store pool is used for this. It can be recovered when all lookup caches are
968 flushed. Lookups now release memory at the end of their result strings.
969 This has involved some general refactoring of the lookup sources.
971 7. Some code has been added to the store_xxx() functions to reduce the amount
972 of flapping under certain conditions.
974 8. log_incoming_interface used to affect only the <= reception log lines. Now
975 it causes the local interface and port to be added to several more SMTP log
976 lines, for example "SMTP connection from", and rejection lines.
978 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
980 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
982 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
983 could be overwritten at the end of the current message (or the start of a
984 new message if it was set in a HELO ACL). The value is now preserved for
985 the duration of the SMTP connection.
987 12. If a transport had a headers_rewrite setting, and a matching header line
988 contained an unqualified address, that address was qualified, even if it
989 did not match any rewriting rules. The underlying bug was that the values
990 of the flags that permit the existence of unqualified sender and recipient
991 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
992 non-local messages, and by -bnq for local messages) were not being
993 preserved with the message after it was received.
995 13. When Exim was logging an SMTP synchronization error, it could sometimes log
996 "next input=" as part of the text comprising the host identity instead of
997 the correct text. The code was using the same buffer for two different
998 strings. However, depending on which order the printing function evaluated
999 its arguments, the bug did not always show up. Under Linux, for example, my
1000 test suite worked just fine.
1002 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1003 doesn't work with some older versions of Perl. It has been changed to "my",
1004 which in any case is probably the better facility to use.
1006 15. A really picky compiler found some instances of statements for creating
1007 error messages that either had too many or two few arguments for the format
1010 16. The size of the buffer for calls to the DNS resolver has been increased
1011 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1012 for addresses that have a lot of PTR records. This alleviates a problem; it
1013 does not fully solve it.
1015 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1016 buffer now truncates the list and logs the incident, which is the same
1017 action as happens when Exim is looking up a host name and its aliases.
1018 Previously in this situation something unpredictable would happen;
1019 sometimes it was "internal error: store_reset failed".
1021 18. If a server dropped the connection unexpectedly when an Exim client was
1022 using GnuTLS and trying to read a response, the client delivery process
1023 crashed while trying to generate an error log message.
1025 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1026 string, for example:
1028 warn message = H1: something\nH2: something
1030 the text was added as a single header line from Exim's point of view
1031 though it ended up OK in the delivered message. However, searching for the
1032 second and subsequent header lines using $h_h2: did not work. This has been
1033 fixed. Similarly, if a system filter added multiple headers in this way,
1034 the routers could not see them.
1036 20. Expanded the error message when iplsearch is called with an invalid key to
1037 suggest using net-iplsearch in a host list.
1039 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1040 ACLs are no longer actually imposed (and a message to that effect is
1043 22. If a "gecos" field in a passwd entry contained escaped characters, in
1044 particular, if it contained a \" sequence, Exim got it wrong when building
1045 a From: or a Sender: header from that name. A second bug also caused
1046 incorrect handling when an unquoted " was present following a character
1047 that needed quoting.
1049 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1050 was not being matched caselessly.
1052 24. Arranged for all hyphens in the exim.8 source to be escaped with
1055 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1056 properly. Recipient callout cache records were still being keyed to include
1057 the sender, even when use_sender was set false. This led to far more
1058 callouts that were necessary. The sender is no longer included in the key
1059 when use_sender is false.
1061 26. Added "control = submission" modifier to ACLs.
1063 27. Added the ${base62d: operator to decode base 62 numbers.
1065 28. dnsdb lookups can now access SRV records.
1067 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1068 the configuration file.
1070 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1071 (-v) mode. This makes the output for a verbose queue run more intelligible.
1073 31. Added a use_postmaster feature to recipient callouts.
1075 32. Added the $body_zerocount variable, containing the number of binary zero
1076 bytes in the message body.
1078 33. The time of last modification of the "new" subdirectory is now used as the
1079 "mailbox time last read" when there is a quota error for a maildir
1082 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1084 35. Added +ignore_unknown as a special item in host lists.
1086 36. Code for decoding IPv6 addresses in host lists is now included, even if
1087 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1088 address was recognized as an IP address, but was then not correctly decoded
1089 into binary, causing unexpected and incorrect effects when compared with
1096 1. Very minor rewording of debugging text in manualroute to say "list of
1097 hosts" instead of "hostlist".
1099 2. If verify=header_syntax was set, and a header line with an unqualified
1100 address (no domain) and a large number of spaces between the end of the
1101 name and the colon was received, the reception process suffered a buffer
1102 overflow, and (when I tested it) crashed. This was caused by some obsolete
1103 code that should have been removed. The fix is to remove it!
1105 3. When running in the test harness, delay a bit after writing a bounce
1106 message to get a bit more predictability in the log output.
1108 4. Added a call to search_tidyup() just before forking a reception process. In
1109 theory, someone could use a lookup in the expansion of smtp_accept_max_
1110 per_host which, without the tidyup, could leave open a database connection.
1112 5. Added the variables $recipient_data and $sender_data which get set from a
1113 lookup success in an ACL "recipients" or "senders" condition, or a router
1114 "senders" option, similar to $domain_data and $local_part_data.
1116 6. Moved the writing of debug_print from before to after the "senders" test
1119 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1120 problems for message scanning, either using a data ACL, or using
1121 local_scan() because the Received: header was not generated till after they
1122 were called (in order to set the time as the time of reception completion).
1123 I have revised the way this works. The header is now generated after the
1124 body is received, but before the ACL or local_scan() are called. After they
1125 are run, the timestamp in the header is updated.
1131 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1132 before starting a queue runner without re-exec. This happened only when
1133 deliver_drop_privilege was set or when the Exim user was set to root. The
1134 effect of the bug was that timeouts during subsequent deliveries caused
1135 crashes instead of being properly handled. The handler is now left at its
1136 default (and expected) setting.
1138 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1139 message, again when deliver_drop_privilege is set or Exim is run as root.
1140 The bug described in (1) was not present in this case, but the tidying up
1141 of the other signals was missing. I have made the two cases consistent.
1143 3. The ignore_target_hosts setting on a manualroute router was being ignored
1144 for hosts that were looked up using the /MX notation.
1146 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1149 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1150 operated on the sender address. After changing the $sender_address to <>
1151 for the sender address verify, Exim was re-instated it as the original
1152 (before rewriting) address, but remembering that it had rewritten it, so it
1153 wasn't rewriting it again. This bug also had the effect of breaking the
1154 sender address verification caching when the sender address was rewritten.
1156 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1157 This has been changed so that if the ip literal address matches
1158 ignore_target_hosts, the router declines.
1160 7. Added expansion conditions match_domain, match_address, and match_local_
1161 part (NOT match_host).
1163 8. The placeholder for the Received: header didn't have a length field set.
1165 9. Added code to Exim itself and to exim_lock to test for a specific race
1166 condition that could lead to file corruption when using MBX delivery. The
1167 issue is with the lockfile that is created in /tmp. If this file is removed
1168 after a process has opened it but before that process has acquired a lock,
1169 there is the potential for a second process to recreate the file and also
1170 acquire a lock. This could lead to two Exim processes writing to the file
1171 at the same time. The added code performs the same test as UW imapd; it
1172 checks after acquiring the lock that its file descriptor still refers to
1173 the same named file.
1175 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1176 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1179 11. Added the smtp_active_hostname option. If used, this will typically be made
1180 to depend on the incoming interface address. Because $interface_address is
1181 not set up until the daemon has forked a reception process, error responses
1182 that can happen earlier (such as "too many connections") no longer contain
1185 12. If an expansion in a condition on a "warn" statement fails because a lookup
1186 defers, the "warn" statement is abandoned, and the next ACL statement is
1187 processed. Previously this caused the whole ACL to be aborted.
1189 13. Added the iplsearch lookup type.
1191 14. Added ident_timeout as a log selector.
1193 15. Added tls_certificate_verified as a log selector.
1195 16. Added a global option tls_require_ciphers (compare the smtp transport
1196 option of the same name). This controls incoming TLS connections.
1198 17. I finally figured out how to make tls_require_ciphers do a similar thing
1199 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1200 before starting the TLS session.
1202 18. Tabs are now shown as \t in -bP output.
1204 19. If the log selector return_path_on_delivery was set, Exim crashed when
1205 bouncing a message because it had too many Received: header lines.
1207 20. If two routers both had headers_remove settings, and the first one included
1208 a superfluous trailing colon, the final name in the first list and the
1209 first name in the second list were incorrectly joined into one item (with a
1210 colon in the middle).
1216 1. Added -C and -D options to the exinext utility, mainly to make it easier
1217 to include in the automated testing, but these could be helpful when
1218 multiple configurations are in use.
1220 2. The exinext utility was not formatting the output nicely when there was
1221 an alternate port involved in the retry record key, nor when there was a
1222 message id as well (for retries that were specific to a specific message
1223 and a specific host). It was also confused by IPv6 addresses, because of
1224 the additional colons they contain. I have fixed the IPv4 problem, and
1225 patched it up to do a reasonable job for IPv6.
1227 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1228 delivery, the log line now contains "pipelined" if PIPELINING was used.
1230 4. An SMTP transport process used to panic and die if the bind() call to set
1231 an explicit outgoing interface failed. This has been changed; it is now
1232 treated in the same way as a connect() failure.
1234 5. A reference to $sender_host_name in the part of a conditional expansion
1235 that was being skipped was still causing a DNS lookup. This no longer
1238 6. The def: expansion condition was not recognizing references to header lines
1239 that used bh_ and bheader_.
1241 7. Added the _cache feature to named lists.
1243 8. The code for checking quota_filecount in the appendfile transport was
1244 allowing one more file than it should have been.
1246 9. For compatibility with Sendmail, the command line option
1254 and sets the incoming protocol and host name (for trusted callers). The
1255 host name and its colon can be omitted when only the protocol is to be set.
1256 Note the Exim already has two private options, -pd and -ps, that refer to
1257 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1258 "s", but I don't think that's a major issue.
1260 10. A number of refactoring changes to the code, none of which should affect
1263 (a) The number of logging options was getting close to filling up the
1264 32-bit word that was used as a bit map. I have split them into two classes:
1265 those that are passed in the argument to log_write(), and those that are
1266 only ever tested independently outside of that function. These are now in
1267 separate 32-bit words, so there is plenty of room for expansion again.
1268 There is no change in the user interface or the logging behaviour.
1270 (b) When building, for example, log lines, the code previously used a
1271 macro that called string_cat() twice, in order to add two strings. This is
1272 not really sufficiently general. Furthermore, there was one instance where
1273 it was actually wrong because one of the argument was used twice, and in
1274 one call a function was used. (As it happened, calling the function twice
1275 did not affect the overall behaviour.) The macro has been replaced by a
1276 function that can join an arbitrary number of extra strings onto a growing
1279 (c) The code for expansion conditions now uses a table and a binary chop
1280 instead of a serial search (which was left over from when there were very
1281 few conditions). Also, it now recognizes conditions like "pam" even when
1282 the relevant support is not compiled in: a suitably worded error message is
1283 given if an attempt is made to use such a condition.
1285 11. Added ${time_interval:xxxxx}.
1287 12. A bug was causing one of the ddress fields not to be passed back correctly
1288 from remote delivery subprocesses. The field in question was not being
1289 subsequently used, so this caused to problems in practice.
1291 13. Added new log selectors queue_time and deliver_time.
1293 14. Might have fixed a bug in maildirsizefile handling that threw up
1294 "unexpected character" debug warnings, and recalculated the data
1295 unnecessarily. In any case, I expanded the warning message to give more
1298 15. Added the message "Restricted characters in address" to the statements in
1299 the default ACL that block characters like @ and % in local parts.
1301 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1302 Three changes have been made:
1304 (a) There was a serious bug; a negative response to MAIL caused the whole
1305 recipient domain to be cached as invalid, thereby blocking all messages
1306 to all local parts at the same domain, from all senders. This bug has
1307 been fixed. The domain is no longer cached after a negative response to
1308 MAIL if the sender used is not empty.
1310 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1313 (c) A new callout option, "use_sender" has been added for people who want
1314 the modified behaviour.
1320 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1323 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1324 indeed breaks things for older releases.
1326 3. Added additional logging to the case where there is a problem reading data
1327 from a filter that is running in a subprocess using a pipe, in order to
1328 try to track down a specific problem.
1330 4. Testing facility fudge: when running in the test harness and attempting
1331 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1332 getting "No route to host". Convert this to a timeout.
1334 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1337 6. Some OS don't have socklen_t but use size_t instead. This affects the
1338 fifth argument of getsockopt() amongst other things. This is now
1339 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1340 can be set for individual OS. I have set it for SunOS5, OSF1, and
1341 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1342 some earlier ones do not.
1344 7. Change 4.30/15 was not doing the test caselessly.
1346 8. The standard form for an IPv6 address literal was being rejected by address
1347 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1348 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1349 this, as well as the form without the "IPv6" on the front (but only when
1350 address literals are enabled, of course).
1352 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1354 10. Exim crashed if a message with an empty sender address specified by -f
1355 encountered a router with an errors_to setting. This could be provoked only
1356 by a command such as
1360 where an empty string was supplied; "<>" did not hit this bug.
1362 11. Installed PCRE release 4.5.
1364 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1365 remained set. It is now erased.
1367 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1368 times from message ids (which are base 36 rather than the normal 62).
1370 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1371 were being counted as actual protocol errors, and logged if the log
1372 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1373 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1374 and DATA following a set of rejected RCPTs do not count as protocol errors.
1375 In other words, Exim assumes they were pipelined, though this may not
1376 actually be the case. Of course, in all cases the client gets an
1377 appropriate error code.
1379 15. If a lookup fails in an ACL condition, a message about the failure may
1380 be available; it is used if testing the ACL cannot continue, because most
1381 such messages specify what the cause of the deferral is. However, some
1382 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1383 that caused an old message to be retained and used if a later statement
1384 caused a defer, replacing the real cause of the deferral.
1386 16. If an IP address had so many PTR records that the DNS lookup buffer
1387 was not large enough to hold them, Exim could crash while trying to process
1388 the truncated data. It now detects and logs this case.
1390 17. Further to 4.21/58, another change has been made: if (and only if) the
1391 first line of a message (the first header line) ends with CRLF, a bare LF
1392 in a subsequent header line has a space inserted after it, so as not to
1393 terminate the header.
1395 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1396 unnecessarily, used atoi() instead of strtol(), and didn't check the
1397 termination when getting file sizes from file names by regex.
1399 19. Completely re-implemented the support for maildirsize files, in the light
1400 of a number of problems with the previous contributed implementation
1401 (4.30/29). In particular:
1403 . If the quota is zero, the maildirsize file is maintained, but no quota is
1406 . If the maildir directory does not exist, it is created before any attempt
1407 to write a maildirsize file.
1409 . The quota value in the file is just a cache; if the quota is changed in
1410 the transport, the new value overrides.
1412 . A regular expression is available for excluding directories from the
1415 20. The autoreply transport checks the characters in options that define the
1416 message's headers; it allows continued headers, but it was checking with
1417 isspace() after an embedded newline instead of explicitly looking for a
1420 21. If all the "regular" hosts to which an address was routed had passed their
1421 expiry times, and had not reached their retry times, the address was
1422 bounced, even if fallback hosts were defined. Now Exim should go on to try
1425 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1426 equivalent code in the SMTP transport. Some hosts send humungous responses
1427 to HELO/EHLO, more than 1024 it seems.
1429 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1430 gives compiler warnings in some environments. I've now done it "properly",
1433 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1434 (because of problems with the built-in one) was declared to return uschar *
1435 instead of char *, causing compiler failure.
1437 25. Fixed a file descriptor leak when processing alias/forward files.
1439 26. Fixed a minor format string issue in dbfn.c.
1441 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1443 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1444 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1447 29. When a sender address is verified, it is cached, to save repeating the test
1448 when there is more than one recipient in a message. However, when the
1449 verification involves a callout, it is possible for different callout
1450 options to be set for different recipients. It is too complicated to keep
1451 track of this in the cache, so now Exim always runs a verification when a
1452 callout is required, relying on the callout cache for the optimization.
1453 The overhead is duplication of the address routing, but this should not be
1456 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1457 to be verified with callout=postmaster, and the main callout worked but the
1458 postmaster check failed, the verification correctly failed. However, if a
1459 subsequent RCPT command asked for sender verification *without* the
1460 postmaster check, incorrect caching caused this verification also to fail,
1463 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1464 it was not caching the DNS options (qualify_single, search_parents) that
1465 were used when the lookup failed. A subsequent lookup with different
1466 options therefore always gave the same answer, though there were cases
1467 where it should not have. (Example: a "domains = !$mx_any" option on a
1468 dnslookup router: the "domains" option is always processed without any
1469 widening, but the router might have qualify_single set.) Now Exim uses the
1470 cached value only when the same options are set.
1472 32. Added John Jetmore's "exipick" utility to the distribution.
1474 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1475 than a timeout (e.g. a certificate is required, and is not provided), an
1476 Exim server now closes the connection immediately. Previously it waited for
1477 the client to close - but if the client is SSL, it seems that they each
1478 wait for each other, leading to a delay before one of them times out.
1480 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1481 maintained 0.8.x compatibility because I don't think many are using it, and
1482 it is clearly obsolete.
1484 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1487 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1488 client certificate was expired. A simple patch fixes this, though I don't
1489 understand the full logic of why the verify callback is called multiple
1492 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1493 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1494 which causes problems with some clients (such as the Certicom SSL Plus
1495 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1496 disables the coutermeasure allowing Eudora to connect."
1498 38. Exim was not checking that a write() to a log file succeeded. This could
1499 lead to Bad Things if a log got too big, in particular if it hit a file
1500 size limit. Exim now panics and dies if it cannot write to a log file, just
1501 as it does if it cannot open a log file.
1503 39. Modified OS/Makefile-Linux so that it now contains
1505 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1507 The two -D definitions ensure that Exim is compiled with large file
1508 support, which makes it possible to handle log files that are bigger than
1511 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1512 instance) a domain was checked against a named list that involved a lookup,
1513 causing $domain_data to be set, then another domain was checked against the
1514 same list, then the first domain was re-checked, the value of $domain_data
1515 after the final check could be wrong. In particular, if the second check
1516 failed, it could be set empty. This bug probably also applied to
1519 41. The strip_trailing_dot option was not being applied to the address given
1520 with the -f command-line option.
1522 42. The code for reading a message's header from the spool was incrementing
1523 $received_count, but never initializing it. This meant that the value was
1524 incorrect (doubled) while delivering a message in the same process in which
1525 it was received. In the most common configuration of Exim, this never
1526 happens - a fresh exec is done - but it can happen when
1527 deliver_drop_privilege is set.
1529 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1530 it now includes up to 150 characters of the unexpected data in the log
1533 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1534 and building data strings. The size of both of these buffers was 10 000
1535 bytes - far larger than anybody would *ever* want, thought I. Needless to
1536 say, somebody hit the limit. I have increased the maximum line length to
1537 20 000 and the maximum data length of concatenated lines to 100 000. I have
1538 also fixed two bugs, because there was no checking on these buffers. Tsk,
1539 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1540 buffer is too small.
1542 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1543 lsearch lookups. Now it does.
1545 46. When parsing a route_list item in a manualroute router, a fixed-length
1546 buffer was used for the list of hosts. I made this 1024 bytes long,
1547 thinking that nobody would ever have a list of hosts that long. Wrong.
1548 Somebody had a whole pile of complicated expansion conditions, and the
1549 string was silently truncated, leading to an expansion error. It turns out
1550 that it is easier to change to an unlimited length (owing to other changes
1551 that have happened since this code was originally written) than to build
1552 structure for giving a limitation error. The length of the item that
1553 expands into the list of hosts is now unlimited.
1555 47. The lsearch lookup could not handle data where the length of text line was
1556 more than 4095 characters. Such lines were truncated, leading to shortened
1557 data being returned. It should now handle lines of any length.
1559 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1560 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1563 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1564 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1565 no longer happens in comment lines. A list of replacements is now placed
1566 at the head of all of the source files, except those whose only change is
1567 to replace PERL_COMMAND in the very first #! line.
1569 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1570 messages on the queue, with a bottom-up merge sort, using code contributed
1571 by Michael Haardt. This should make operations like -bp somewhat faster on
1572 large queues. It won't affect queue runners, except when queue_run_in_order
1575 51. Installed eximstats 1.31 in the distribution.
1577 52. Added support for SRV lookups to the dnslookup router.
1579 53. If an ACL referred to $message_body or $message_body_end, the value was not
1580 reset for any messages that followed in the same SMTP session.
1582 54. The store-handling optimization for building very long strings was not
1583 differentiating between the different store pools. I don't think this
1584 actually made any difference in practice, but I've tidied it.
1586 55. While running the routers to verify a sender address, $sender_address
1587 was still set to the sender address. This is wrong, because when routing to
1588 send a bounce to the sender, it would be empty. Therefore, I have changed
1589 it so that, while verifying a sender address, $sender_address is set to <>.
1590 (There is no change to what happens when verifying a recipient address.)
1592 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1593 target A or AAAA records (if not already returned) without resetting the
1594 qualify_single or search_parents options of the DNS resolver. These are
1595 inappropriate in this case because the targets of MX and SRV records must
1596 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1597 target that, when qualified, matched something in the local domain. These
1598 two options are now turned off when doing these lookups.
1600 57. It seems that at least some releases of Reiserfs (which does not have the
1601 concept of a fixed number of inodes) returns zero and not -1 for the
1602 number of available inodes. This interacted badly with check_spool_inodes,
1603 which assumed that -1 was the "no such thing" setting. What I have done is
1604 to check that the total number of inodes is greater than zero before doing
1605 the test of how many are available.
1607 58. When a "warn" ACL statement has a log_message modifier, the message is
1608 remembered, and not repeated. This is to avoid a lot of repetition when a
1609 message has many recipients that cause the same warning to be written.
1610 Howewer, Exim was preserving the list of already written lines for an
1611 entire SMTP session, which doesn't seem right. The memory is now reset if a
1612 new message is started.
1614 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1615 debugging output unless log_rewrite was also set.
1617 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1618 of (int)(handle) when we know that handle contains (void *)(-1).
1620 61. The Exim daemon panic-logs an error return when it closes the incoming
1621 connection. However "connection reset by peer" seems to be common, and
1622 isn't really an error worthy of noting specially, so that particular error
1625 62. When Exim is trying to find all the local interfaces, it used to panic and
1626 die if the ioctl to get the interface flags failed. However, it seems that
1627 on at least one OS (Solaris 9) it is possible to have an interface that is
1628 included in the list of interfaces, but for which you get a failure error
1629 for this call. This happens when the interface is not "plumbed" into a
1630 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1631 failure of the "get flags" call assumes that the interface is down.
1633 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1634 makes life easier for people who are doing arithmetic on fields extracted
1635 from dates, where you often get leading zeros that should not be
1636 interpreted as octal.
1638 64. Added qualify_domain to the redirect router, to override the global
1641 65. If a pathologically long header line contained very many addresses (the
1642 report of this problem mentioned 10 000) and each of them was rewritten,
1643 Exim could use up a very large amount of memory. (It kept on making new
1644 copies of the header line as it rewrote, and never released the old ones.)
1645 At the expense of a bit more processing, the header rewriting function has
1646 been changed so that it no longer eats memory in this way.
1648 66. The generation of the Received: header has been moved from the time that a
1649 message starts to be received, to the time that it finishes. The timestamp
1650 in the Received: header should now be very close to that of the <= log
1651 line. There are two side-effects of this change:
1653 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1654 logged header lines no longer include the local Received: line, because
1655 it has not yet been created. The same applies to a copy of the message
1656 that is returned to a non-SMTP sender when a message is rejected.
1658 (b) When a filter file is tested using -bf, no additional Received: header
1659 is added to the test message. After some thought, I decided that this
1662 This change does not affect the value of $received_for. It is still set
1663 after address rewriting, but before local_scan() is called.
1665 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1667 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1668 gave an unhelpful panic error message, and a defer error. I have managed to
1669 change this behaviour so that it now rejects any supplied certificate,
1670 which seems right, as the list of acceptable certificates is empty.
1672 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1673 gave an unhelpful defer error. I have not managed to make this reject any
1674 supplied certificates, but the error message it gives is "no certificate
1675 supplied", which is not helpful.
1677 70. exigrep's output now also includes lines that are not associated with any
1678 message, but which match the given pattern. Implemented by a patch from
1679 Martin Sluka, which also tidied up the Perl a bit.
1681 71. Recipient callout verification, like sender verification, was using <> in
1682 the MAIL FROM command. This isn't really the right thing, since the actual
1683 sender may affect whether the remote host accepts the recipient or not. I
1684 have changed it to use the actual sender in the callout; this means that
1685 the cache record is now keyed on a recipient/sender pair, not just the
1686 recipient address. There doesn't seem to be a real danger of callout loops,
1687 since a callout by the remote host to check the sender would use <>.
1688 [SEE ABOVE: changed after hitting problems.]
1690 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1691 temporary errors. However, in the case of such a code being given after
1692 the end of a data transmission (i.e. after ".") Exim was failing to write
1693 a retry record for the message. (Yes, there was some broken host that was
1694 actually sending 8xx at this point.)
1696 73. An unknown lookup type in a host list could cause Exim to panic-die when
1697 the list was checked. (An example that provoked this was putting <; in the
1698 middle of a list instead of at the start.) If this happened during a DATA
1699 ACL check, a -D file could be left lying around. This kind of configuration
1700 error no longer causes Exim to die; instead it causes a defer errror. The
1701 incident is still logged to the main and panic logs.
1703 74. Buglet left over from Exim 3 conversion. The message "too many messages
1704 in one connection" was written to the rejectlog but not the mainlog, except
1705 when address rewriting (yes!) was being logged.
1707 75. Added write_rejectlog option.
1709 76. When a system filter was run not as root (that is, when system_filter_user
1710 was set), the values of the $n variables were not being returned to the
1711 main process; thus, they were not subsequently available in the $sn
1714 77. Added +return_path_on_delivery log selector.
1716 78. A connection timeout was being treated differently from recipients deferred
1717 when testing hosts_max_try with a message that was older than the host's
1718 retry timeout. (The host should not be counted, thus allowing all hosts to
1719 be tried at least once before bouncing.) This may have been the cause of an
1720 occasionally reported bug whereby a message would remain on the queue
1721 longer than the retry timeout, but would be bounced if a delivery was
1722 forced. I say "may" because I never totally pinned down the problem;
1723 setting up timeout/retry tests is difficult. See also the next item.
1725 79. The ultimate address timeout was not being applied to errors that involved
1726 a combination of host plus message (for example, a timeout on a MAIL
1727 command). When an address resolved to a number of possible hosts, and they
1728 were not all tried for each delivery (e.g. because of hosts_max_try), a
1729 message could remain on the queue longer than the retry timeout.
1731 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1734 81. Fixed an obscure SMTP outgoing bug which required at least the following
1735 conditions: (a) there was another message waiting for the same server;
1736 (b) the server returned 5xx to all RCPT commands in the first message so
1737 that the message was not completed; (c) the server dropped the connection
1738 or gave a negative response to the RSET that Exim sends to abort the
1739 transaction. The observed case was a dropped connection after DATA that had
1740 been sent in pipelining mode. That is, the server had advertised PIPELINING
1741 but was not implementing it correctly. The effect of the bug was incorrect
1742 behaviour, such as trying another host, and this could lead to a crash.
1748 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1749 and daemon.c were passed as pointers to ints; they should have been
1750 pointers to socklen_t variables (which are typically unsigned ints).
1752 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1755 3. Fixed a really odd bug that affected only the testing scheme; patching a
1756 certain fixed string in the binary changed the value of another string that
1757 happened to be identical to the end of the original first string.
1759 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1760 name", it returns that address as the IP address. On some operating
1761 systems (e.g. Solaris), it also passes back the IP address string as the
1762 "host name". However, on others (e.g. Linux), it passes back an empty
1763 string. Exim wasn't checking for this, and was changing the host name to an
1764 empty string, assuming it had been canonicized.
1766 5. Although rare, it is permitted to have more than one PTR record for a given
1767 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1768 all the names associated with an address, because they do in Solaris.
1769 However, it seems that they do not in Linux for data that comes from the
1770 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1771 I found this out when I moved to a new Linux workstation and tried to run
1772 the Exim test suite.
1774 To get round this problem I have changed the code so that it now does its
1775 own call to the DNS to look up PTR records when searching for a host name.
1776 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1777 addresses that are only in /etc/hosts are still found.
1779 This behaviour is, however, controlled by an option called host_lookup_
1780 order, which defaults to "bydns:byaddr". If people want to use the other
1781 order, or indeed, just use one or the other means of lookup, they can
1782 specify it in this variable.
1784 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1785 some operating systems, this comes back from gethostbyaddr() as an empty
1786 string, and this is what Exim used to test for. However, it seems that in
1787 other systems, "." is yielded. Exim now tests for this case too.
1789 7. The values of check_spool_space and check_log_space are now held internally
1790 as a number of kilobytes instead of an absolute number of bytes. If a
1791 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1792 kilobyte. This means that much larger values can be stored.
1794 8. Exim monitor: an attempt to get the action menu when not actually pointing
1795 at a message produces an empty menu entitled "No message selected". This
1796 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1797 no entries in it ("Shell widget menu has zero width and/or height"). So I
1798 have added a single, blank menu entry in this case.
1800 9. Added ${quote_local_part.
1802 10. MIME decoding is now applied to the contents of Subject: header lines when
1805 11. Now that a reference to $sender_host_address automatically causes a reverse
1806 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1807 host lookup before query-style lookups in lists that might use this
1808 variable. This has therefore been abolished, and the "net-" prefix is no
1809 longer necessary for query-style lookups.
1811 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1812 have been a typo for LFLAGS, so it has been changed.
1814 13. The install script calls Exim with "-C /dev/null" in order to find the
1815 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1816 to be output. Howeve, since Exim outputs its version number before the
1817 error, it didn't break the script. It just looked ugly. I fixed this by
1818 always allowing "-C /dev/null" if the caller is root.
1820 14. Ignore overlarge ACL variable number when reading spool file - insurance
1821 against a later release with more variables having written the file.
1823 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1824 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1825 this, as well as the form without the "IPv6" on the front.
1827 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1828 OS/Makefile-Darwin file.
1830 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1831 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1833 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1834 variable. However, some OS use ints for inodes. I've added cast to long int
1835 to get rid of the compiler warning.
1837 19. I had forgotten to lock out "/../" in configuration file names when
1838 ALT_CONFIG_PREFIX was set.
1840 20. Routers used for verification do not need to specify transports. However,
1841 if such a router generated a host list, and callout was configured, Exim
1842 crashed, because it could not find a port number from the (non-existent)
1843 transport. It now assumes port 25 in this circumstance.
1845 21. Added the -t option to exigrep.
1847 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1848 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1849 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1850 them is set, LOOKUP_LSEARCH is forced.
1852 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1854 24. Added sender and host information to the "rejected by local_scan()" log
1855 line; previously there was no indication of these.
1857 25. Added .include_if_exists.
1859 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1860 Linux. It turns out that not all file systems support this. Apparently some
1861 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1862 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1863 not supported on the file descriptor, is now ignored when Exim is trying to
1864 sync a directory. This applies only to Linux.
1866 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1868 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1869 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1872 29. Added support for maildirsize files from supplied patch (modified a bit).
1874 30. The use of :fail: followed by an empty string could lead Exim to respond to
1875 sender verification failures with (e.g.):
1877 550 Verification failed for <xxx>
1878 550 Sender verify failed
1880 where the first response line was missing the '-' that indicates it is not
1881 the final line of the response.
1883 31. The loop for finding the name of the user that called Exim had a hardwired
1884 limit of 10; it now uses the value of finduser_retries, which is used for
1885 all other user lookups.
1887 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1890 33. Exim was neglecting to zero errno before one call of strtol() when
1891 expanding a string and expecting an integer value. On some systems this
1892 resulted in spurious "integer overflow" errors. Also, it was casting the
1893 result into an int without checking.
1895 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1896 did not work. The code looks as if it has *never* worked, though it appears
1897 to have been documented since at least releast 1.62. I have made it work.
1899 35. The "timeout_DNS" error in retry rules, also documented since at least
1900 1.62, also never worked. As it isn't clear exactly what this means, and
1901 clearly it isn't a major issue, I have abolished the feature by treating it
1902 as "timeout", and writing a warning to the main and panic logs.
1904 36. The display of retry rules for -brt wasn't always showing the error code
1907 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1908 timeout_connect_A, timeout_connect_MX.
1910 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1911 to the empty sender.
1913 39. The daemon was not analysing the content of -oX till after it had closed
1914 stderr and disconnected from the controlling terminal. This meant that any
1915 syntax errors were only noted on the panic log, and the return code from
1916 the command was 0. By re-arranging the code a little, I've made the
1917 decoding happen first, so such errors now appear on stderr, and the return
1918 code is 1. However, the actual setting up of the sockets still happens in
1919 the disconnected process, so errors there are still only recorded on the
1922 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1923 connections (as happens on some IP stacks) was logged at start up time as
1924 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1925 it from "IPv6 and IPv4", which means that two separate sockets are being
1928 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1929 says whether AF_INET or AF_INET6 was passed as an argument.
1931 42. Exiwhat output was messed up when time zones were included in log
1934 43. Exiwhat now gives more information about the daemon's listening ports,
1935 and whether -tls-on-connect was used.
1937 44. The "port" option of the smtp transport is now expanded.
1939 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1940 silently ignored. Now an error message is written to the main and panic
1943 46. There's a new ACL modifier called "logwrite" which writes to a log file
1944 as soon as it is encountered.
1946 47. Added $local_user_uid and $local_user_gid at routing time.
1948 48. Exim crashed when trying to verify a sender address that was being
1951 49. Exim was recognizing only a space character after ".include". It now also
1952 recognizes a tab character.
1954 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1955 extracting the relevant information from the specification. The man page no
1956 longer contains scrambled data for the -d option, and I've added a section
1957 at the front about calling Exim under different names.
1959 51. Added "extra_headers" argument to the "mail" command in filter files.
1961 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1964 53. Installed eximstats 1.29.
1966 54. Added transport_filter_timeout as a generic transport option.
1968 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1969 Cc: header lines. This was required by RFC 822, but it not required by RFC
1972 56. Exim used to add From:, Date:, and Message-Id: header lines to any
1973 incoming messages that did not have them. Now it does so only if the
1974 message originates locally, that is, if there is no associated remote host
1975 address. When Resent- header lines are present, this applies to the Resent-
1976 lines rather than the non-Resent- lines.
1978 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
1979 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
1981 58. Messages for configuration errors now include the name of the main
1982 configuration file - useful now that there may be more than one file in a
1983 list (.included file names were always shown).
1985 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1986 for those rare installations that do not start the daemon as root or run it
1987 setuid root. I've cut out the call to initgroups() if the daemon is not
1990 60. The Exim user and group can now be bound into the binary as text strings
1991 that are looked up at the start of Exim's processing.
1993 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
1995 62. Added $mailstore_basename variable.
1997 63. Installed patch to sieve.c from Michael Haardt.
1999 64. When Exim failed to open the panic log after failing to open the main log,
2000 the original message it was trying to log was written to stderr and debug
2001 output, but if they were not available (the usual case in production), it
2002 was lost. Now it is written to syslog before the two lines that record the
2003 failures to open the logs.
2005 65. Users' Exim filters run in subprocesses under the user's uid. It is
2006 possible for a "deliver" command or an alias in a "personal" command to
2007 provoke an address rewrite. If logging of address rewriting is configured,
2008 this fails because the process is not running as root or exim. There may be
2009 a better way of dealing with this, but for the moment (because 4.30 needs
2010 to be released), I have disabled address rewrite logging when running a
2011 filter in a non-root, non-exim process.
2017 1. The buildconfig auxiliary program wasn't quoting the value set for
2018 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2019 not defined. This bug was masked in 4.22 by the effect that was fixed in
2022 2. Some messages that were rejected after a message id was allocated were
2023 shown as "incomplete" by exigrep. It no longer does this for messages that
2024 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2026 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2027 have allow_domain_literals set, the ID did not get logged in the <= line.
2028 Domain literals are now always recognized in Message-ID: header lines.
2030 4. The first argument for a ${extract expansion item is the key name or field
2031 number. Leading and trailing spaces in this item were not being ignored,
2032 causing some misleading effects.
2034 5. When deliver_drop_privilege was set, single queue runner processes started
2035 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2036 same command in the process it spins off) were not dropping privilege.
2038 6. When the daemon running as "exim" started a queue runner, it always
2039 re-executed Exim in the spun-off process. This is a waste of effort when
2040 deliver_drop_privilege is set. The new process now just calls the
2041 queue-runner function directly.
2047 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2050 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2051 ignored. Though the use of -oP was forcing the writing of a pid file, it
2052 was always written to the default place.
2054 3. If the message "no IP address found for host xxxx" is generated during
2055 incoming verification, it is now followed by identification of the incoming
2056 connection (so you can more easily find what provoked it).
2058 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2060 5. Added some features to "harden" Exim a bit more against certain attacks:
2062 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2063 be put in Local/Makefile. This is like the never_users runtime option,
2064 but it cannot be overridden. The default setting is "root".
2066 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2067 prefix string with which any file named in a -C command line option
2070 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2071 is retained for -C and -D only if the caller of Exim is root. Without
2072 it, the exim user may also use -C and -D and retain privilege.
2074 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2075 command line option is disabled.
2077 6. Macro names set by the -D option must start with an upper case letter, just
2078 like macro names defined in the configuration file.
2080 7. Added "dereference=" facility to LDAP.
2082 8. Two instances of the typo "uknown" in the source files are fixed.
2084 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2085 the Configure-Makefile script screwed up while processing it.
2087 10. Incorporated PCRE 4.4.
2089 11. The SMTP synchronization check was not operating right at the start of an
2090 SMTP session. For example, it could not catch a HELO sent before the client
2091 waited for the greeting. There is now a check for outstanding input at the
2092 point when the greeting is written. Because of the duplex, asynchronous
2093 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2094 way, but not yet received, when the check is performed.
2096 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2097 on TCP/IP sockets, because this apparently causes some broken clients to
2100 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2101 unchanged) from the Cygwin maintainer.
2103 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2104 was supported instead of testing for mbx. Effectively a typo.
2106 15. The spa authenticator server code was not checking that the input it
2107 received was valid base64.
2109 16. The debug output line for the "set" modifier in ACLs was not showing the
2110 name of the variable that was being set.
2112 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2114 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2115 lookup on its own. Something else was needed to trigger the lookup. For
2116 example, a match in host_lookup or the need for a host name in a host list.
2117 Now, if $sender_host_name is referenced and the host name has not yet been
2118 looked up, a lookup is performed. If the lookup fails, the variable remains
2119 empty, and $host_lookup_failed is set to "1".
2121 19. Added "eqi" as a case-independent comparison operator.
2123 20. The saslauthd authentication condition could segfault if neither service
2124 nor realm was specified.
2126 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2127 error message that was logged was misleading, and incoming SMTP
2128 connections were dropped. The message is now more accurate, and temporary
2129 errors are given to SMTP connections.
2131 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2132 (except RSET) with a 421 error, until QUIT is received. However, it was
2133 failing to send a response to QUIT.
2135 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2136 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2137 ACL is now run after the helo_try_verify_hosts code.
2139 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2140 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2141 case-independent, so other case variants are also recognized). Apparently
2142 some systems use these upper case variants.
2144 25. If more than two messages were waiting for the same host, and a transport
2145 filter was specified for the transport, Exim sent two messages over the
2146 same TCP/IP connection, and then failed with "socket operation on non-
2147 socket" when it tried to send the third.
2149 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2151 27. The extern definition of crypt16() in expand.c was not being excluded when
2152 the OS had its own crypt16() function.
2154 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2155 as a preferred synonym for return_size_limit, both as an option and as an
2158 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2160 30. Changed the default configuration ACL to relax the local part checking rule
2161 for addresses that are not in any local domains. For these addresses,
2162 slashes and pipe symbols are allowed within local parts, but the sequence
2163 /../ is explicitly forbidden.
2165 31. SPA server authentication was not clearing the challenge buffer before
2168 32. log_message in a "warn" ACL statement was writing to the reject log as
2169 well as to the main log, which contradicts the documentation and doesn't
2170 seem right (because no rejection is happening). So I have stopped it.
2172 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2173 However, I am unable to do any testing of this.
2175 34. Fixed an infelicity in the appendfile transport. When checking directories
2176 for a mailbox, to see if any needed to be created, it was accidentally
2177 using path names with one or more superfluous leading slashes; tracing
2178 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2180 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2181 contents are added to the log line that is written for every discarded
2182 recipient. (Previously a log_message setting was ignored.)
2184 36. The ${quote: operator now quotes the string if it is empty.
2186 37. The install script runs exim in order to find its version number. If for
2187 some reason other than non-existence or emptiness, which it checks, it
2188 could not run './exim', it was installing it with an empty version number,
2189 i.e. as "exim-". This error state is now caught, and the installation is
2192 38. An argument was missing from the function that creates an error message
2193 when Exim fails to connect to the socket for saslauthd authentication.
2194 This could cause Exim to crash, or give a corrupted message.
2196 39. Added isip, isip4, and isip6 to ${if conditions.
2198 40. The ACL variables $acl_xx are now saved with the message, and can be
2199 accessed later in routers, transports, and filters.
2201 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2202 strings in the file are not string-expanded.
2204 42. If a MAIL command specified a SIZE value that was too large to fit into an
2205 int variable, the check against message_size_limit failed. Such values are
2206 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2207 day this will have to be increased, but I don't think I want to be around
2208 when emails are that large.
2215 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2216 iconv() is not standard in FreeBSD.
2218 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2219 IPv6 enabled. The observed symptom was a segmentation fault on return from
2220 the function os_common_find_running_interfaces() in src/os.c.
2222 3. In the check_special_case() function in daemon.c I had used "errno" as an
2223 argument name, which causes warnings on some systems. This was basically a
2224 typo, since it was named "eno" in the comments!
2226 4. The code that waits for the clock to tick (at a resolution of some fraction
2227 of a second) so as to ensure message-id uniqueness was always waiting for
2228 at least one whole tick, when it could have waited for less. [This is
2229 almost certainly not relevant at current processor speeds, where it is
2230 unlikely to ever wait at all. But we try to future-proof.]
2232 5. The function that sleeps for a time interval that includes fractions of a
2233 second contained a race. It did not block SIGALRM between setting the
2234 timer, and suspending (a couple of lines later). If the interval was short
2235 and the sigsuspend() was delayed until after it had expired, the suspension
2236 never ended. On busy systems this could lead to processes getting stuck for
2239 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2240 process, before it forks any delivery processes. The open lookup caching
2241 mechanism meant that the open file or database connection was passed into
2242 the delivery process. The problem was that delivery processes always tidy
2243 up cached lookup data. This could cause a problem for the next delivery
2244 process started by the queue runner, because the external queue runner
2245 process does not know about the closure. So the next delivery process
2246 still has data in the lookup cache. In the case of a file lookup, there was
2247 no problem because closing a file descriptor in a subprocess doesn't affect
2248 the parent. However, if the lookup was caching a connection to a database,
2249 the connection was closed, and the second delivery process was likely to
2250 see errors such as "PGSQL: query failed: server closed the connection
2251 unexpectedly". The problem has been fixed by closing all cached lookups
2252 in a queue runner before running a delivery process.
2254 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2255 seem to have the "const" qualifier which it has on other OS. I've
2258 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2259 *of the same type* (client or server) with the same public name that an
2260 error should be diagnosed.
2262 9. When Exim looked up a host name for an IP address, but failed to find the
2263 original IP address when looking up the host name (a safety check), it
2264 output the message "<ip address> does not match any IP for NULL", which was
2265 confusing, to say the least. The bug was that the host name should have
2266 appeared instead of "NULL".
2268 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2269 user that is built into the binary, and the -C or -D options is used, root
2270 privilege is dropped before the configuration file is read. In addition,
2271 logging is switched to stderr instead of the normal log files. If the
2272 configuration then re-defines the Exim user, the unprivileged environment
2273 is probably not what is expected, so Exim logs a panic warning message (but
2276 However, if deliver_drop_privilege is set, the unprivileged state may well
2277 be exactly what is intended, so the warning has been cut out in that case,
2278 and Exim is allowed to try to write to its normal log files.
2284 1. smtp_return_error_details was not giving details for temporary sender
2285 or receiver verification errors.
2287 2. Diagnose a configuration error if two authenticators have the same public
2290 3. Exim used not to create the message log file for a message until the first
2291 delivery attempt. This could be confusing when incoming messages were held
2292 for policy or load reasons. The message log file is now created at the time
2293 the message is received, and an initial "Received" line is written to it.
2295 4. The automatically generated man page for command line options had a minor
2296 bug that caused no ill effects; however, a more serious problem was that
2297 the procedure for building the man page automatically didn't always
2298 operate. Consequently, release 4.20 contains an out-of-date version. This
2299 shouldn't happen again.
2301 5. When building Exim with embedded Perl support, the script that builds the
2302 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2303 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2305 6. The freeze_tell option was not being used for messages that were frozen on
2306 arrival, either by an ACL or by local_scan().
2308 7. Added the smtp_incomplete_transaction log selector.
2310 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2311 was accepting AUTH without a new EHLO.
2313 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2314 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2317 10. Logging of TCP/IP connections (when configured) now happens in the main
2318 daemon process instead of the child process, so that the TCP/IP connection
2319 count is more accurate (but it can never be perfect).
2321 11. The use of "drop" in a nested ACL was not being handled correctly in the
2322 outer ACL. Now, if condition failure induced by the nested "drop" causes
2323 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2324 or "require"), the connection is dropped.
2326 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2327 that yield "discard" can now be used with an "accept" or a "discard" verb,
2328 but an error is generated for any others (because I can't see a useful way
2329 to define what should happen).
2331 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2332 are now processed in the same way as lines in the Exim configuration file.
2333 In particular, continuation lines are supported.
2335 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2337 15. Added -ti meaning -t -i.
2339 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2340 domains, and warn by logging if others are found.
2342 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2343 returned from GIFCONF when Exim is trying to find the list of interfaces on
2344 a host. The code in os.c has been modified to copy each ifreq to an aligned
2345 structure in all cases.
2347 Also, in some cases, the returned ifreq's were being copied to a 'struct
2348 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2349 means the last couple of bytes of an IPv6 address could be chopped if the
2350 ifreq contained only a normal sockaddr (14 bytes storage).
2352 18. Named domain lists were not supported in the hosts_treat_as_local option.
2353 An entry such as +xxxx was not recognized, and was treated as a literal
2356 19. Ensure that header lines added by a DATA ACL are included in the reject log
2357 if the ACL subsequently rejects the message.
2359 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2360 MD5 (which is deprecated).
2362 21. When testing a filter file using -bf, Exim was writing a message when it
2363 took the sender from a "From " line in the message, but it was not doing so
2364 when it took $return_path from a Return-Path: header line. It now does.
2366 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2367 with a valid header line field name (a series of printing characters
2368 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2370 23. Changed "disc" in the source to "disk" to conform to the documentation and
2371 the book and for uniformity.
2373 24. Ignore Sendmail's -Ooption=value command line item.
2375 25. When execve() failed while trying to run a command in a pipe transport,
2376 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2377 could be confused with a return value of 69 from the command itself. This
2378 has been changed to 127, the value the shell returns if it is asked to run
2379 a non-existent command. The wording for the related log line suggests a
2380 non-existent command as the problem.
2382 26. If received_header_text expands to an empty string, do not add a Received:
2383 header line to the message. (Well, it adds a token one on the spool, but
2384 marks it "old" so that it doesn't get used or transmitted.)
2386 27. Installed eximstats 1.28 (addition of -nt option).
2388 28. There was no check for failure on the call to getsockname() in the daemon
2389 code. This can fail if there is a shortage of resources on the system, with
2390 ENOMEM, for example. A temporary error is now given on failure.
2392 29. Contrary to the C standard, it seems that in some environments, the
2393 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2394 program. Exim now does this explicitly; it affects the formatting of
2395 timestamps using strftime().
2397 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2398 complaints. I've now initialized all the variables, to avoid this.
2400 32. Header lines added by a system filter were not being "seen" during
2401 transport-time rewrites.
2403 33. The info_callback() function passed to OpenSSL is set up with type void
2404 (*)(SSL *, int, int), as described somewhere. However, when calling the
2405 function (actually a macro) that sets it up, the type void(*)() is
2406 expected. I've put in a cast to prevent warnings from picky compilers.
2408 34. If a DNS black list lookup found a CNAME record, but there were no A
2409 records associated with the domain it pointed at, Exim crashed.
2411 35. If a DNS black list lookup returned more than one A record, Exim ignored
2412 all but the first. It now scans all returned addresses if a particular IP
2413 value is being sought. In this situation, the contents of the
2414 $dnslist_value variable are a list of all the addresses, separated by a
2417 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2418 to accept a host name and all its aliases if the forward lookup for any of
2419 them yielded the IP address of the incoming connection. Now it accepts only
2420 those names whose forward lookup yields the correct IP address. Any other
2421 names are discarded. This closes a loophole whereby a rogue DNS
2422 administrator could create reverse DNS records to break through a
2423 wildcarded host restriction in an ACL.
2425 37. If a user filter or a system filter that ran in a subprocess used any of
2426 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2427 the wrong values were passed to the pipe command ($thisaddress had the
2428 value of $0, $0 had the value of $1, etc). This bug was introduced by
2429 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2431 38. Improved the line breaking for long SMTP error messages from ACLs.
2432 Previously, if there was no break point between 40 and 75 characters, Exim
2433 left the rest of the message alone. Two changes have been made: (a) I've
2434 reduced the minimum length to 35 characters; (b) if it can't find a break
2435 point between 35 and 75 characters, it looks ahead and uses the first one
2436 that it finds. This may give the occasional overlong line, but at least the
2437 remaining text gets split now.
2439 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2440 file descriptors might be low, and that setting 1000 would always raise it.
2441 It turns out that in some environments, the limit is already over 1000 and
2442 that lowering it causes trouble. So now Exim takes care not to decrease it.
2444 40. When delivering a message, the value of $return_path is set to $sender_
2445 address at the start of routing (routers may change the value). By an
2446 oversight, this default was not being set up when an address was tested by
2447 -bt or -bv, which affected the outcome if any router or filter referred to
2450 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2451 log only when "message" or "log_message" are set. However, if one of the
2452 conditions was an address verification, or a call to a nested ACL, the
2453 messages generated by the underlying test were being passed through. This
2454 no longer happens. The underlying message is available in $acl_verify_
2455 message for both "message" and "log_message" expansions, so it can be
2456 passed through if needed.
2458 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2459 new expansion $bh_ to give the encoded byte string without charset
2460 translation. Translation happens only if iconv() is available; HAVE_ICONV
2461 indicates this at build time. HEADERS_CHARSET gives the charset to
2462 translate to; headers_charset can change it in the configuration, and
2463 "headers charset" can change it in an individual filter file.
2465 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2466 that creates RFC 2047 encoded "words" labels them as that charset instead
2467 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2468 expansion operator; (ii) when Exim creates a From: line for a local
2469 message; (iii) when a header line is rewritten to include a "phrase" part.
2471 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2472 buggy, causing it to skip the first lines of messages whose message ID
2473 ended in 'D'. This would not have bitten before Exim release 4.14, because
2474 message IDs were unlikely to end in 'D' before then. The effect was to have
2475 incorrect size information for certain domains.
2477 45. #include "config.h" was missing at the start of the crypt16.c module. This
2478 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2481 46. If there was a timeout during a "random" callout check, Exim treated it as
2482 a failure of the random address, and carried on sending RSET and the real
2483 address. If the delay was just some slowness somewhere, the response to the
2484 original RCPT would be taken as a response to RSET and so on, causing
2485 mayhem of various kinds.
2487 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2488 when I implemented it. It didn't allow for the fact that some option values
2489 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2490 the right test for positive values.
2492 48. Domain names in DNS records are case-independent. Exim always looks them up
2493 in lower case. Some resolvers return domain names in exactly the case they
2494 appear in the zone file, that is, they may contain uppercase letters. Not
2495 all resolvers do this - some return always lower case. Exim was treating a
2496 change of case by a resolver as a change of domain, similar to a widening
2497 of a domain abbreviation. This triggered its re-routing code and so it was
2498 trying to route what was effectively the same domain again. This normally
2499 caused routing to fail (because the router wouldn't handle the domain
2500 twice). Now Exim checks for this case specially, and just changes the
2501 casing of the domain that it ultimately uses when it transmits the message
2504 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2507 50. If a filter generated a file delivery with a non-absolute name (possible if
2508 no home directory exists for the router), the forbid_file option was not
2511 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2512 the existing equality matching.
2514 52. Exim was using ints instead of ino_t variables in some places where it was
2515 dealing with inode numbers.
2517 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2518 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2519 called TMPDIR, and if it finds it is different, it changes its value.
2521 54. The smtp_printf() function is now made available to local_scan() so
2522 additional output lines can be written before returning. There is also an
2523 smtp_fflush() function to enable the detection of a dropped connection.
2524 The variables smtp_input and smtp_batched_input are exported to
2527 55. Changed the default runtime configuration: the message "Unknown user"
2528 has been removed from the ACL, and instead placed on the localuser router,
2529 using the cannot_route_message feature. This means that any verification
2530 failures that generate their own messages won't get overridden. Similarly,
2531 the "Unrouteable address" message that was in the ACL for unverifiable
2532 relay addresses has also been removed.
2534 56. Added hosts_avoid_esmtp to the smtp transport.
2536 57. The exicyclog script was not checking for the esoteric option
2537 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2538 will work only if exicyclog is run under the appropriate euid.
2540 58. Following a discussion on the list, the rules by which Exim recognises line
2541 endings on incoming messages have been changed. The -dropcr and drop_cr
2542 options are now no-ops, retained only for backwards compatibility. The
2543 following line terminators are recognized: LF CRLF CR. However, special
2544 processing applies to CR:
2546 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2547 nor a local message in the state where . is a terminator.
2549 (ii) If a bare CR is encountered in a header line, an extra space is added
2550 after the line terminator so as not to end the header. The reasoning
2551 behind this is that bare CRs in header lines are most likely either
2552 to be mistakes, or people trying to play silly games.
2554 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2555 was being incorrectly given as 18 bytes larger than it should have been.
2556 This is a VOB (very old bug).
2558 60. This may never have affected anything current, but just in case it has:
2559 When the local host is found other than at the start of a list of hosts,
2560 the local host, those with the same MX, and any that follow, are discarded.
2561 When the list in question was part of a longer list of hosts, the following
2562 hosts (not currently being processed) were also being discarded. This no
2563 longer happens. I'm not sure if this situation could ever has previously
2566 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2569 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2570 header. This is something that is recommended in a new Internet Draft, and
2571 is something that is documented as being done by Sendmail. There are two
2572 possible values. For messages generated by the autoreply transport, Exim
2575 Auto-Submitted: auto-replied
2577 whereas for all other generated messages (e.g. bounces) it adds
2579 Auto-Submitted: auto-generated
2581 63. The "personal" condition in filters now includes a test for the
2582 Auto-Submitted: header. If it contains the string "auto-" the message it
2583 not considered personal.
2585 64. Added rcpt_include_affixes as a generic transport option.
2587 65. Added queue_only_override (default true).
2589 66. Added the syslog_duplication option.
2591 67. If what should have been the first header line of a message consisted of
2592 a space followed by a colon, Exim was mis-interpreting it as a header line.
2593 It isn't of course - it is syntactically invalid and should therefore be
2594 treated as the start of the message body. The misbehaviour could have
2595 caused a number of strange effects, including loss of data in subsequent
2596 header lines, and spool format errors.
2598 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2599 client host had authenticated. This control can now be exercised by an ACL
2600 for more flexibility.
2602 69. By default, callouts do not happen when testing with -bh. There is now a
2603 variant, -bhc, which does actually run the callout code, including
2604 consulting and updating the callout cache.
2606 70. Added support for saslauthd authentication, courtesy of Alexander
2609 71. If statvfs() failed on the spool or log directories while checking their
2610 size for availability, Exim confusingly gave the error "space shortage".
2611 Furthermore, in debugging mode it crashed with a floating point exception.
2612 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2613 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2614 serious problem, Exim now writes to the main and panic logs when this
2615 happens, with details of the failure. It then refuses to accept the
2616 incoming message, giving the message "spool directory problem" or "log
2617 directory problem" with a 421 code for SMTP messages.
2619 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2620 0, 1, and 2 exist, because some OS complain for execs without them (see
2621 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2622 descriptors. However, the code omitted to check that the open succeeded,
2623 causing mysterious errors if for some reason the permissions on /dev/null
2624 got screwed. Now Exim writes a message to the main and panic logs, and
2625 bombs out if it can't open /dev/null.
2627 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2628 interact so that it is all more flexible. It is supposed to remain
2629 backwards compatible. Also added extra_local_interfaces.
2631 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2632 to bomb out with an assertion failure - to the client this appears as a
2633 connection drop. This problem occurs in the part of the code that was taken
2634 from the Samba project. Fortunately, the assertion is in a very simple
2635 function, so I have fixed this by reproducing the function inline in the
2636 one place where it is called, and arranging for authentication to fail
2637 instead of killing the process with assert().
2639 75. The SPA client code was not working when the server requested OEM rather
2640 than Unicode encoding.
2642 76. Added code to make require_files with a specific uid setting more usable in
2643 the case where statting the file as root fails - usually a non-root-mounted
2644 NFS file system. When this happens and the failure is EACCES, Exim now
2645 forks a subprocess and does the per-uid checking as the relevant uid.
2647 77. Added process_log_path.
2649 78. If log_file_path was not explicitly set, a setting of check_log_space or
2650 check_log_inodes was ignored.
2652 79. If a space check for the spool or log partitions fails, the incident is now
2653 logged. Of course, in the latter case the data may get lost...
2655 80. Added the %p formatting code to string_format() so that it can be used to
2656 print addresses in debug_print(). Adjusted all the address printing in the
2657 debugging in store.c to use %p rather than %d.
2659 81. There was a concern that a line of code in smtp_in.c could overflow a
2660 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2661 initially expressed, the concern was not well-founded, because trailing
2662 spaces are removed early. However, if the trailing spaces were followed by
2663 a NULL, they did not get removed, so the overflow was possible. Two fixes
2666 (a) I re-wrote the offending code in a cleaner fashion.
2667 (b) If an incoming SMTP command contains a NULL character, it is rejected
2670 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2671 runs initgroups(), so that if the Exim user is in any additional groups,
2672 they will be used during message reception.
2678 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob