1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.88 2005/03/09 14:41:00 tom 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
29 A note about Exim versions 4.44 and 4.50
30 ----------------------------------------
32 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
33 changes of various kinds. As a consequence, a big documentation update was
34 needed. This delayed the release for rather longer than seemed good, especially
35 in the light of a couple of (minor) security issues. Therefore, the changes
36 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
37 release. So 4.44 and 4.50 are in effect two different branches that both start
40 I have left the 4.50 change log unchanged; it contains all the changes since
41 4.43. The change log for 4.44 is below; many of its items are identical to
42 those for 4.50. This seems to be the most sensible way to preserve the
43 historical information.
49 1. Minor wording change to the doc/README.SIEVE file.
51 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
52 computation of the current number of files was incorrect.
54 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
55 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
58 4. Give more explanation in the error message when the command for a transport
59 filter fails to execute.
61 5. There are several places where Exim runs a non-Exim command in a
62 subprocess. The SIGUSR1 signal should be disabled for these processes. This
63 was being done only for the command run by the queryprogram router. It is
64 now done for all such subprocesses. The other cases are: ${run, transport
65 filters, and the commands run by the lmtp and pipe transports.
67 6. Added CONFIGURE_GROUP build-time option.
69 7. Some older OS have a limit of 256 on the maximum number of file
70 descriptors. Exim was using setrlimit() to set 1000 as a large value
71 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
72 systems. I've change it so that if it can't get 1000, it tries for 256.
74 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
75 was an oversight, and furthermore, ever since the addition of extra
76 controls (e.g. 4.43/32), the checks on when to allow different forms of
77 "control" were broken. There should now be diagnostics for all cases when a
78 control that does not make sense is encountered.
80 9. Added the /retain_sender option to "control=submission".
82 10. $recipients is now available in the predata ACL (oversight).
84 11. Tidy the search cache before the fork to do a delivery from a message
85 received from the command line. Otherwise the child will trigger a lookup
86 failure and thereby defer the delivery if it tries to use (for example) a
87 cached ldap connection that the parent has called unbind on.
89 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
90 of $address_data from the recipient verification was clobbered by the
93 13. The value of address_data from a sender verification is now available in
94 $sender_address_data in subsequent conditions in the ACL statement.
96 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
98 15. Added a new option "connect=<time>" to callout options, to set a different
101 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
102 was its contents. (It was OK if the option was not defined at all.)
104 17. A "Completed" log line is now written for messages that are removed from
105 the spool by the -Mrm option.
107 18. New variables $sender_verify_failure and $recipient_verify_failure contain
108 information about exactly what failed.
110 19. Added -dd to debug only the daemon process.
112 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
113 handles timeouts, both on the server side and network timeouts. Renamed the
114 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
116 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
117 It is now set to "smtps".
119 22. $host_address is now set to the target address during the checking of
122 23. When checking ignore_target_hosts for an ipliteral router, no host name was
123 being passed; this would have caused $sender_host_name to have been used if
124 matching the list had actually called for a host name (not very likely,
125 since this list is usually IP addresses). A host name is now passed as
128 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
129 code that specifies a non-restarting handler (typically sigaction() in
130 modern systems) in an attempt to fix a rare and obscure crash bug.
132 25. Narrowed the window for a race in the daemon that could cause it to ignore
133 SIGCHLD signals. This is not a major problem, because they are used only to
134 wake it up if nothing else does.
136 26. A malformed maildirsize file could cause Exim to calculate negative values
137 for the mailbox size or file count. Odd effects could occur as a result.
138 The maildirsize information is now recalculated if the size or filecount
141 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
142 support for a long time. Removed HAVE_SYS_VFS_H.
144 28. Installed the latest version of exipick from John Jetmore.
146 29. In an address list, if the pattern was not a regular expression, an empty
147 subject address (from a bounce message) matched only if the pattern was an
148 empty string. Non-empty patterns were not even tested. This was the wrong
149 because it is perfectly reasonable to use an empty address as part of a
150 database query. An empty address is now tested by patterns that are
151 lookups. However, all the other forms of pattern expect the subject to
152 contain a local part and a domain, and therefore, for them, an empty
153 address still always fails if the pattern is not itself empty.
155 30. Exim went into a mad DNS loop when attempting to do a callout where the
156 host was specified on an smtp transport, and looking it up yielded more
159 31. Re-factored the code for checking spool and log partition space into a
160 function that finds that data and another that does the check. The former
161 is then used to implement four new variables: $spool_space, $log_space,
162 $spool_inodes, and $log_inodes.
164 32. The RFC2047 encoding function was originally intended for short strings
165 such as real names; it was not keeping to the 75-character limit for
166 encoded words that the RFC imposes. It now respects the limit, and
167 generates multiple encoded words if necessary. To be on the safe side, I
168 have increased the buffer size for the ${rfc2047: expansion operator from
171 33. It is now permitted to omit both strings after an "if" condition; if the
172 condition is true, the result is "true". As before, when the second string
173 is omitted, a false condition yields an empty string. This makes it less
174 cumbersome to write custom ACL and router conditions.
176 34. Failure to deliver a bounce message always caused it to be frozen, even if
177 there was an errors_to setting on the router. The errors_to setting is now
180 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
181 canonical form (fully expanded) before being placed in
182 $sender_host_address.
184 36. The table in the code that translates DNS record types into text (T_A to
185 "A" for instance) was missing entries for NS and CNAME. It is just possible
186 that this could have caused confusion if both these types were looked up
187 for the same domain, because the text type is used as part of Exim's
188 per-process caching. But the chance of anyone hitting this buglet seems
191 37. The dnsdb lookup has been extended in a number of ways.
193 (1) There is a new type, "zns", which walks up the domain tree until it
194 finds some nameserver records. It should be used with care.
196 (2) There is a new type, "mxh", which is like "mx" except that it returns
197 just the host names, not the priorities.
199 (3) It is now possible to give a list of domains (or IP addresses) to be
200 looked up. The behaviour when one of the lookups defers can be
201 controlled by a keyword.
203 (4) It is now possible to specify the separator character for use when
204 multiple records are returned.
206 38. The dnslists ACL condition has been extended: it is now possible to supply
207 a list of IP addresses and/or domains to be looked up in a particular DNS
210 39. Added log_selector=+queue_time_overall.
212 40. When running the queue in the test harness, wait just a tad after forking a
213 delivery process, to get repeatability of debugging output.
215 41. Include certificate and key file names in error message when GnuTLS fails
216 to set them up, because the GnuTLS error message doesn't include the name
217 of the failing file when there is a problem reading it.
219 42. Allow both -bf and -bF in the same test run.
221 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
223 44. The "Exiscan patch" is now merged into the mainline Exim source.
225 45. Sometimes the final signoff response after QUIT could fail to get
226 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
227 < 0 before doing a fflush(). This bug looks as though it goes back to the
228 introduction of TLS in release 3.20, but "sometimes" must have been rare
229 because the tests only now provoked it.
231 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
232 (this can affect the format of dates).
234 47. exim_tidydb, when checking for the continued existence of a message for
235 which it has found a message-specific retry record, was not finding
236 messages that were in split spool directories. Consequently, it was
237 deleting retry records that should have stayed in existence.
239 48. Steve fixed some bugs in eximstats.
241 49. The SPA authentication driver was not abandoning authentication and moving
242 on to the next authenticator when an expansion was forced to fail,
243 contradicting the general specification for all authenticators. Instead it
244 was generating a temporary error. It now behaves as specified.
246 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
247 (the order specifies the preference for clients). The order is now AES256,
248 AES128, 3DES, ARCFOUR128.
250 51. Small patch to Sieve code - explicitly set From: when generating an
253 52. Exim crashed if a remote delivery caused a very long error message to be
254 recorded - for instance if somebody sent an entire SpamAssassin report back
255 as a large number of 550 error lines. This bug was coincidentally fixed by
256 increasing the size of one of Exim's internal buffers (big_buffer) that
257 happened as part of the Exiscan merge. However, to be on the safe side, I
258 have made the code more robust (and fixed the comments that describe what
261 53. Now that there can be additional text after "Completed" in log lines (if
262 the queue_time_overall log selector is set), a one-byte patch to exigrep
263 was needed to allow it to recognize "Completed" as not the last thing in
266 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
267 patch that reportedly fixes this has been added. I am not expert enough to
268 create a test for it. This is what the patch creator wrote:
270 "I found a little strange behaviour of ldap code when working with
271 Windows 2003 AD Domain, where users was placed in more than one
272 Organization Units. When I tried to give exim partial DN, the exit code
273 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
274 But simultaneously result of request was absolutely normal ldap result,
275 so I produce this patch..."
277 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
278 REFERENCE, so I have modified the code to exclude the patch when that macro
281 55. Some experimental protocols are using DNS PTR records for new purposes. The
282 keys for these records are domain names, not reversed IP addresses. The
283 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
284 leaves it alone. Component reversal etc. now happens only for IP addresses.
287 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
289 57. Double the size of the debug message buffer (to 2048) so that more of very
290 long debug lines gets shown.
292 58. The exicyclog utility now does better if the number of log files to keep
293 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
295 59. Two changes related to the smtp_active_hostname option:
297 (1) $smtp_active_hostname is now available as a variable.
298 (2) The default for smtp_banner uses $smtp_active_hostname instead
299 of $primary_hostname.
301 60. The host_aton() function is supposed to be passed a string that is known
302 to be a valid IP address. However, in the case of IPv6 addresses, it was
303 not checking this. This is a hostage to fortune. Exim now panics and dies
304 if the condition is not met. A case was found where this could be provoked
305 from a dnsdb PTR lookup with an IPv6 address that had more than 8
306 components; fortuitously, this particular loophole had already been fixed
307 by change 4.50/55 above.
309 If there are any other similar loopholes, the new check in host_aton()
310 itself should stop them being exploited. The report I received stated that
311 data on the command line could provoke the exploit when Exim was running as
312 exim, but did not say which command line option was involved. All I could
313 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
317 61. There was a buffer overflow vulnerability in the SPA authentication code
318 (which came originally from the Samba project). I have added a test to the
319 spa_base64_to_bits() function which I hope fixes it.
322 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
323 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
325 63. The daemon start-up calls getloadavg() while still root for those OS that
326 need the first call to be done as root, but it missed one case: when
327 deliver_queue_load_max is set with deliver_drop_privilege. This is
328 necessary for the benefit of the queue runner, because there is no re-exec
329 when deliver_drop_privilege is set.
331 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
334 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
335 was in use, was not putting the data itself into the right store pool;
336 consequently, it could be overwritten for a subsequent message in the same
337 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
340 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
342 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
343 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
344 the function were treating the return as a boolean value, which happened to
345 work because 0=false and not-0=true, but is not correct code.
347 68. The host_aton() function was not handling scoped IPv6 addresses (those
348 with, for example, "%eth0" on the end) correctly.
350 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
351 negated items (that is, ~something) in unsigned ints. Some compilers
352 apparently mutter when there is no cast.
354 70. If an address verification called from an ACL failed, and did not produce a
355 user-specific message (i.e. there was only a "system" message), nothing was
356 put in $acl_verify_message. In this situation, it now puts the system
359 71. Change 4.23/11 added synchronization checking at the start of an SMTP
360 session; change 4.31/43 added the unwanted input to the log line - except
361 that it did not do this in the start of session case. It now does.
363 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
364 This is wrong and can cause the other end to generate a synchronization
365 error if it is another Exim or anything else that does the synchronization
366 check. A QUIT command is no longer sent after a timeout.
368 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
371 74. The defer_ok option of callout verification was not working if it was used
372 when verifying addresses in header lines, that is, for this case:
374 verify = header_sender/callout=defer_ok
376 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
377 those file descriptors could be used for SMTP connections. If anything
378 wrote to stderr (the example that came up was "warn" in embedded Perl), it
379 could be sent to the SMTP client, causing chaos. The daemon now opens
380 stdin, stdout, and stderr to /dev/null when it puts itself into the
383 76. Arrange for output from Perl's "warn" command to be written to Exim's main
384 log by default. The user can override this with suitable Perl magic.
386 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
387 the log message when discard triggers, was not working for the DATA ACL or
388 for the non-SMTP ACL.
390 78. Error message wording change in sieve.c.
392 79. If smtp_accept_max_per_host was set, the number of connections could be
393 restricted to fewer than expected, because the daemon was trying to set up
394 a new connection before checking whether the processes handling previous
395 connections had finished. The check for completed processes is now done
396 earlier. On busy systems, this bug wouldn't be noticed because something
397 else would have woken the daemon, and it would have reaped the completed
400 80. If a message was submitted locally by a user whose login name contained one
401 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
402 It caused a spool format error. I have fixed the spool reading code. A
403 related problem was that the "from" clause in the Received: line became
404 illegal because of the space(s). It is now covered by ${quote_local_part.
406 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
409 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
411 83. Patch from Sieve maintainer to fix unterminated string problem in
414 84. Some minor changes to the Linux configuration files to help with other
415 OS variants using glibc.
417 85. One more patch for Sieve to update vacation handling to latest spec.
420 ----------------------------------------------------
421 See the note above about the 4.44 and 4.50 releases.
422 ----------------------------------------------------
428 1. Change 4.43/35 introduced a bug that caused file counts to be
429 incorrectly computed when quota_filecount was set in an appendfile
432 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
433 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
436 3. Give more explanation in the error message when the command for a transport
437 filter fails to execute.
439 4. There are several places where Exim runs a non-Exim command in a
440 subprocess. The SIGUSR1 signal should be disabled for these processes. This
441 was being done only for the command run by the queryprogram router. It is
442 now done for all such subprocesses. The other cases are: ${run, transport
443 filters, and the commands run by the lmtp and pipe transports.
445 5. Some older OS have a limit of 256 on the maximum number of file
446 descriptors. Exim was using setrlimit() to set 1000 as a large value
447 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
448 systems. I've change it so that if it can't get 1000, it tries for 256.
450 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
451 was an oversight, and furthermore, ever since the addition of extra
452 controls (e.g. 4.43/32), the checks on when to allow different forms of
453 "control" were broken. There should now be diagnostics for all cases when a
454 control that does not make sense is encountered.
456 7. $recipients is now available in the predata ACL (oversight).
458 8. Tidy the search cache before the fork to do a delivery from a message
459 received from the command line. Otherwise the child will trigger a lookup
460 failure and thereby defer the delivery if it tries to use (for example) a
461 cached ldap connection that the parent has called unbind on.
463 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
464 of $address_data from the recipient verification was clobbered by the
467 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
468 was its contents. (It was OK if the option was not defined at all.)
470 11. A "Completed" log line is now written for messages that are removed from
471 the spool by the -Mrm option.
473 12. $host_address is now set to the target address during the checking of
476 13. When checking ignore_target_hosts for an ipliteral router, no host name was
477 being passed; this would have caused $sender_host_name to have been used if
478 matching the list had actually called for a host name (not very likely,
479 since this list is usually IP addresses). A host name is now passed as
482 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
483 code that specifies a non-restarting handler (typically sigaction() in
484 modern systems) in an attempt to fix a rare and obscure crash bug.
486 15. Narrowed the window for a race in the daemon that could cause it to ignore
487 SIGCHLD signals. This is not a major problem, because they are used only to
488 wake it up if nothing else does.
490 16. A malformed maildirsize file could cause Exim to calculate negative values
491 for the mailbox size or file count. Odd effects could occur as a result.
492 The maildirsize information is now recalculated if the size or filecount
495 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
496 support for a long time. Removed HAVE_SYS_VFS_H.
498 18. Updated exipick to current release from John Jetmore.
500 19. Allow an empty sender to be matched against a lookup in an address list.
501 Previously the only cases considered were a regular expression, or an
504 20. Exim went into a mad DNS lookup loop when doing a callout where the
505 host was specified on the transport, if the DNS lookup yielded more than
508 21. The RFC2047 encoding function was originally intended for short strings
509 such as real names; it was not keeping to the 75-character limit for
510 encoded words that the RFC imposes. It now respects the limit, and
511 generates multiple encoded words if necessary. To be on the safe side, I
512 have increased the buffer size for the ${rfc2047: expansion operator from
515 22. Failure to deliver a bounce message always caused it to be frozen, even if
516 there was an errors_to setting on the router. The errors_to setting is now
519 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
520 canonical form (fully expanded) before being placed in
521 $sender_host_address.
523 24. Updated eximstats to version 1.33
525 25. Include certificate and key file names in error message when GnuTLS fails
526 to set them up, because the GnuTLS error message doesn't include the name
527 of the failing file when there is a problem reading it.
529 26. Expand error message when OpenSSL has problems setting up cert/key files.
532 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
533 (this can affect the format of dates).
535 28. exim_tidydb, when checking for the continued existence of a message for
536 which it has found a message-specific retry record, was not finding
537 messages that were in split spool directories. Consequently, it was
538 deleting retry records that should have stayed in existence.
540 29. eximstats updated to version 1.35
541 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
542 1.35 - bugfix such that pie charts by volume are generated correctly
544 30. The SPA authentication driver was not abandoning authentication and moving
545 on to the next authenticator when an expansion was forced to fail,
546 contradicting the general specification for all authenticators. Instead it
547 was generating a temporary error. It now behaves as specified.
549 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
550 (the order specifies the preference for clients). The order is now AES256,
551 AES128, 3DES, ARCFOUR128.
553 31. Small patch to Sieve code - explicitly set From: when generating an
556 32. Exim crashed if a remote delivery caused a very long error message to be
557 recorded - for instance if somebody sent an entire SpamAssassin report back
558 as a large number of 550 error lines. This bug was coincidentally fixed by
559 increasing the size of one of Exim's internal buffers (big_buffer) that
560 happened as part of the Exiscan merge. However, to be on the safe side, I
561 have made the code more robust (and fixed the comments that describe what
564 33. Some experimental protocols are using DNS PTR records for new purposes. The
565 keys for these records are domain names, not reversed IP addresses. The
566 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
567 leaves it alone. Component reversal etc. now happens only for IP addresses.
570 34. The host_aton() function is supposed to be passed a string that is known
571 to be a valid IP address. However, in the case of IPv6 addresses, it was
572 not checking this. This is a hostage to fortune. Exim now panics and dies
573 if the condition is not met. A case was found where this could be provoked
574 from a dnsdb PTR lookup with an IPv6 address that had more than 8
575 components; fortuitously, this particular loophole had already been fixed
576 by change 4.50/55 or 4.44/33 above.
578 If there are any other similar loopholes, the new check in host_aton()
579 itself should stop them being exploited. The report I received stated that
580 data on the command line could provoke the exploit when Exim was running as
581 exim, but did not say which command line option was involved. All I could
582 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
586 35. There was a buffer overflow vulnerability in the SPA authentication code
587 (which came originally from the Samba project). I have added a test to the
588 spa_base64_to_bits() function which I hope fixes it.
591 36. The daemon start-up calls getloadavg() while still root for those OS that
592 need the first call to be done as root, but it missed one case: when
593 deliver_queue_load_max is set with deliver_drop_privilege. This is
594 necessary for the benefit of the queue runner, because there is no re-exec
595 when deliver_drop_privilege is set.
597 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
598 was in use, was not putting the data itself into the right store pool;
599 consequently, it could be overwritten for a subsequent message in the same
600 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
603 38. Sometimes the final signoff response after QUIT could fail to get
604 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
605 < 0 before doing a fflush(). This bug looks as though it goes back to the
606 introduction of TLS in release 3.20, but "sometimes" must have been rare
607 because the tests only now provoked it.
613 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
614 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
615 Now it returns an integer. A number of calls were still expecting a T/F
616 return. Fortuitously, in all cases, the tests worked in OK situations,
617 which is the norm. However, things would have gone wrong on any write
618 failures on the smtp file descriptor. This function is used when sending
619 messages over SMTP and also when doing verify callouts.
621 2. When Exim is called to do synchronous delivery of a locally submitted
622 message (the -odf or -odi options), it no longer closes stderr before doing
625 3. Implemented the mua_wrapper option.
627 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
629 5. Implemented the functions header_remove(), header_testname(),
630 header_add_at_position(), and receive_remove_recipient(), and exported them
633 6. If an ACL "warn" statement specified the addition of headers, Exim already
634 inserted X-ACL-Warn: at the start if there was no header name. However, it
635 was not making this test for the second and subsequent header lines if
636 there were newlines in the string. This meant that an invalid header could
637 be inserted if Exim was badly configured.
639 7. Allow an ACL "warn" statement to add header lines at the start or after all
640 the Received: headers, as well as at the end.
642 8. Added the rcpt_4xx retry error code.
644 9. Added postmaster_mailfrom=xxx to callout verification option.
646 10. Added mailfrom=xxxx to the callout verification option, for verify=
649 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
650 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
652 12. Inserted some casts to stop certain compilers warning when using pointer
653 differences as field lengths or precisions in printf-type calls (mostly
654 affecting debugging statements).
656 13. Added optional readline() support for -be (dynamically loaded).
658 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
659 same clock tick as a message's arrival, so that its received time was the
660 same as the "first fail" time on the retry record, and that message
661 remained on the queue past the ultimate address timeout, every queue runner
662 would try a delivery (because it was past the ultimate address timeout) but
663 after another failure, the ultimate address timeout, which should have then
664 bounced the address, did not kick in. This was a "< instead of <=" error;
665 in most cases the first failure would have been in the next clock tick
666 after the received time, and all would be well.
668 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
669 being recognized when the domain list was tested by the match_domain
670 condition in an expansion string.
672 16. Added the ${str2b64: operator.
674 17. Exim was always calling setrlimit() to set a large limit for the number of
675 processes, without checking whether the existing limit was already
676 adequate. (It did check for the limit on file descriptors.) Furthermore,
677 errors from getrlimit() and setrlimit() were being ignored. Now they are
678 logged to the main and panic logs, but Exim does carry on, to try to do its
679 job under whatever limits there are.
681 18. Imported PCRE 5.0.
683 19. Trivial typo in log message " temporarily refused connection" (the leading
686 20. If the log selector return_path_on_delivery was set and an address was
687 redirected to /dev/null, the delivery process crashed because it assumed
688 that a return path would always be set for a "successful" delivery. In this
689 case, the whole delivery is bypassed as an optimization, and therefore no
692 21. Internal re-arrangement: the function for sending a challenge and reading
693 a response while authentication was assuming a zero-terminated challenge
694 string. It's now changed to take a pointer and a length, to allow for
695 binary data in such strings.
697 22. Added the cyrus_sasl authenticator (code supplied by MBM).
699 23. Exim was not respecting finduser_retries when seeking the login of the
700 uid under which it was called; it was always trying 10 times. (The default
701 setting of finduser_retries is zero.) Also, it was sleeping after the final
702 failure, which is pointless.
704 24. Implemented tls_on_connect_ports.
706 25. Implemented acl_smtp_predata.
708 26. If the domain in control=submission is set empty, Exim assumes that the
709 authenticated id is a complete email address when it generates From: or
710 Sender: header lines.
712 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
713 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
714 chown and chgrp in /bin and hostname in /usr/bin.
716 28. Exim was keeping the "process log" file open after each use, just as it
717 does for the main log. This opens the possibility of it remaining open for
718 long periods when the USR1 signal hits a daemon. Occasional processlog
719 errors were reported, that could have been caused by this. Anyway, it seems
720 much more sensible not to leave this file open at all, so that is what now
723 29. The long-running daemon process does not normally write to the log once it
724 has entered its main loop, and it closes the log before doing so. This is
725 so that log files can straightforwardly be renamed and moved. However,
726 there are a couple of unusual error situations where the daemon does write
727 log entries, and I had neglected to close the log afterwards.
729 30. The text of an SMTP error response that was received during a remote
730 delivery was being truncated at 512 bytes. This is too short for some of
731 the long messages that one sometimes sees. I've increased the limit to
734 31. It is now possible to make retry rules that apply only when a message has a
735 specific sender, in particular, an empty sender.
737 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
738 it possible to be selective about when SMTP synchronization is enforced.
740 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
742 32. Implemented hosts_connection_nolog.
744 33. Added an ACL for QUIT.
746 34. Setting "delay_warning=" to disable warnings was not working; it gave a
749 35. Added mailbox_size and mailbox_filecount to appendfile.
751 36. Added control = no_multiline_responses to ACLs.
753 37. There was a bug in the logic of the code that waits for the clock to tick
754 in the case where the clock went backwards by a substantial amount such
755 that the microsecond fraction of "now" was more than the microsecond
756 fraction of "then" (but the whole seconds number was less).
758 38. Added support for the libradius Radius client library this is found on
759 FreeBSD (previously only the radiusclient library was supported).
765 1. When certain lookups returned multiple values in the form name=value, the
766 quoting of the values was not always being done properly. Specifically:
767 (a) If the value started with a double quote, but contained no whitespace,
769 (b) If the value contained whitespace other than a space character (i.e.
770 tabs or newlines or carriage returns) it was not quoted.
771 This fix has been applied to the mysql and pgsql lookups by writing a
772 separate quoting function and calling it from the lookup code. The fix
773 should probably also be applied to nisplus, ibase and oracle lookups, but
774 since I cannot test any of those, I have not disturbed their existing code.
776 2. A hit in the callout cache for a specific address caused a log line with no
777 reason for rejecting RCPT. Now it says "Previous (cached) callout
778 verification failure".
780 3. There was an off-by-one bug in the queryprogram router. An over-long
781 return line was truncated at 256 instead of 255 characters, thereby
782 overflowing its buffer with the terminating zero. As well as fixing this, I
783 have increased the buffer size to 1024 (and made a note to document this).
785 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
786 when Exim is waiting for an SMTP response from a remote server, Exim
787 restarts its select() call on the socket, thereby resetting its timeout.
788 This is not a problem when such interrupts are rare. Somebody set up a cron
789 job to run exiwhat every 2 minutes, which is less than the normal select()
790 timeout (5 or 10 minutes). This meant that the select() timeout never
791 kicked in because it was always reset. I have fixed this by comparing the
792 time when an interrupt arrives with the time at the start of the first call
793 to select(). If more time than the timeout has elapsed, the interrupt is
794 treated as a timeout.
796 5. Some internal re-factoring in preparation for the addition of Sieve
797 extensions (by MH). In particular, the "personal" test is moved to a
798 separate function, and given an option for scanning Cc: and Bcc: (which is
799 not set for Exim filters).
801 6. When Exim created an email address using the login of the caller as the
802 local part (e.g. when creating a From: or Sender: header line), it was not
803 quoting the local part when it contained special characters such as @.
805 7. Installed new OpenBSD configuration files.
807 8. Reworded some messages for syntax errors in "and" and "or" conditions to
808 try to make them clearer.
810 9. Callout options, other than the timeout value, were being ignored when
811 verifying sender addresses in header lines. For example, when using
813 verify = header_sender/callout=no_cache
815 the cache was (incorrectly) being used.
817 10. Added a missing instance of ${EXE} to the exim_install script; this affects
818 only the Cygwin environment.
820 11. When return_path_on_delivery was set as a log selector, if different remote
821 addresses in the same message used different return paths and parallel
822 remote delivery occurred, the wrong values would sometimes be logged.
823 (Whenever a remote delivery process finished, the return path value from
824 the most recently started remote delivery process was logged.)
826 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
827 lines when AUTH and/or TLS are in use. This is the "received protocol"
828 field. Exim used to use "asmtp" for authenticated SMTP, without any
829 indication (in the protocol name) for TLS use. Now it follows the RFC and
830 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
831 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
832 names appear in log lines as well as in Received: header lines.
834 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
835 extensions, and comparison tests, and to fix some bugs.
837 14. Changes to the "personal" filter test:
839 (1) The test was buggy in that it was just doing the equivalent of
840 "contains" tests on header lines. For example, if a user's address was
841 anne@some.where, the "personal" test would incorrectly be true for
843 To: susanne@some.where
845 This test is now done by extracting each address from the header in turn,
846 and checking the entire address. Other tests that are part of "personal"
847 are now done using regular expressions (for example, to check local parts
848 of addresses in From: header lines).
850 (2) The list of non-personal local parts in From: addresses has been
851 extended to include "listserv", "majordomo", "*-request", and "owner-*",
852 taken from the Sieve specification recommendations.
854 (3) If the message contains any header line starting with "List-" it is
855 treated as non-personal.
857 (4) The test for "circular" in the Subject: header line has been removed
858 because it now seems ill-conceived.
860 15. Minor typos in src/EDITME comments corrected.
862 16. Installed latest exipick from John Jetmore.
864 17. If headers_add on a router specified a text string that was too long for
865 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
866 of string_sprintf() is now avoided.
868 18. $message_body_size was not set (it was always zero) when running the DATA
869 ACL and the local_scan() function.
871 19. For the "mail" command in an Exim filter, no default was being set for
872 the once_repeat time, causing a random time value to be used if "once" was
873 specified. (If the value happened to be <= 0, no repeat happened.) The
874 default is now 0s, meaning "never repeat". The "vacation" command was OK
875 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
876 (I found it when inspecting the code).
878 20. There is now an overall timeout for performing a callout verification. It
879 defaults to 4 times the callout timeout, which applies to individual SMTP
880 commands during the callout. The overall timeout applies when there is more
881 than one host that can be tried. The timeout is checked before trying the
882 next host. This prevents very long delays if there are a large number of
883 hosts and all are timing out (e.g. when the network connections are timing
884 out). The value of the overall timeout can be changed by specifying an
885 additional sub-option for "callout", called "maxwait". For example:
887 verify = sender/callout=5s,maxwait=20s
889 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
890 to the end before writing, but this should make it even safer).
892 22. Exim was forgetting that it had advertised PIPELINING for the second and
893 subsequent messages on an SMTP connection. It was also not resetting its
894 memory on STARTTLS and an internal HELO.
896 23. When Exim logs an SMTP synchronization error within a session, it now
897 records whether PIPELINING has been advertised or not.
899 24. Added 3 instances of "(long int)" casts to time_t variables that were being
900 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
901 rather than long int.
903 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
905 26. Added the never_mail option to autoreply.
911 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
912 crash if the getsockname() call failed; this can happen if a connection is
913 closed very soon after it is established. The problem was simply in the
914 order in which certain operations were done, causing Exim to try to write
915 to the SMTP stream before it had set up the file descriptor. The bug has
916 been fixed by making things happen in the correct order.
922 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
923 before the connection was closed, thus losing the rejection response.
925 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
926 some early Solaris releases, but causes trouble in current releases where
927 socklen_t is defined.
929 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
932 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
935 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
936 that releasing the top of it at the end releases what was used for sub-
937 expansions (unless the block got too big). However, discard this block if
938 the first thing is a variable or header, so that we can use its block when
939 it is dynamic (useful for very large $message_headers, for example).
941 6. Lookups now cache *every* query, not just the most recent. A new, separate
942 store pool is used for this. It can be recovered when all lookup caches are
943 flushed. Lookups now release memory at the end of their result strings.
944 This has involved some general refactoring of the lookup sources.
946 7. Some code has been added to the store_xxx() functions to reduce the amount
947 of flapping under certain conditions.
949 8. log_incoming_interface used to affect only the <= reception log lines. Now
950 it causes the local interface and port to be added to several more SMTP log
951 lines, for example "SMTP connection from", and rejection lines.
953 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
955 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
957 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
958 could be overwritten at the end of the current message (or the start of a
959 new message if it was set in a HELO ACL). The value is now preserved for
960 the duration of the SMTP connection.
962 12. If a transport had a headers_rewrite setting, and a matching header line
963 contained an unqualified address, that address was qualified, even if it
964 did not match any rewriting rules. The underlying bug was that the values
965 of the flags that permit the existence of unqualified sender and recipient
966 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
967 non-local messages, and by -bnq for local messages) were not being
968 preserved with the message after it was received.
970 13. When Exim was logging an SMTP synchronization error, it could sometimes log
971 "next input=" as part of the text comprising the host identity instead of
972 the correct text. The code was using the same buffer for two different
973 strings. However, depending on which order the printing function evaluated
974 its arguments, the bug did not always show up. Under Linux, for example, my
975 test suite worked just fine.
977 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
978 doesn't work with some older versions of Perl. It has been changed to "my",
979 which in any case is probably the better facility to use.
981 15. A really picky compiler found some instances of statements for creating
982 error messages that either had too many or two few arguments for the format
985 16. The size of the buffer for calls to the DNS resolver has been increased
986 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
987 for addresses that have a lot of PTR records. This alleviates a problem; it
988 does not fully solve it.
990 17. A dnsdb lookup for PTR records that receives more data than will fit in the
991 buffer now truncates the list and logs the incident, which is the same
992 action as happens when Exim is looking up a host name and its aliases.
993 Previously in this situation something unpredictable would happen;
994 sometimes it was "internal error: store_reset failed".
996 18. If a server dropped the connection unexpectedly when an Exim client was
997 using GnuTLS and trying to read a response, the client delivery process
998 crashed while trying to generate an error log message.
1000 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1001 string, for example:
1003 warn message = H1: something\nH2: something
1005 the text was added as a single header line from Exim's point of view
1006 though it ended up OK in the delivered message. However, searching for the
1007 second and subsequent header lines using $h_h2: did not work. This has been
1008 fixed. Similarly, if a system filter added multiple headers in this way,
1009 the routers could not see them.
1011 20. Expanded the error message when iplsearch is called with an invalid key to
1012 suggest using net-iplsearch in a host list.
1014 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1015 ACLs are no longer actually imposed (and a message to that effect is
1018 22. If a "gecos" field in a passwd entry contained escaped characters, in
1019 particular, if it contained a \" sequence, Exim got it wrong when building
1020 a From: or a Sender: header from that name. A second bug also caused
1021 incorrect handling when an unquoted " was present following a character
1022 that needed quoting.
1024 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1025 was not being matched caselessly.
1027 24. Arranged for all hyphens in the exim.8 source to be escaped with
1030 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1031 properly. Recipient callout cache records were still being keyed to include
1032 the sender, even when use_sender was set false. This led to far more
1033 callouts that were necessary. The sender is no longer included in the key
1034 when use_sender is false.
1036 26. Added "control = submission" modifier to ACLs.
1038 27. Added the ${base62d: operator to decode base 62 numbers.
1040 28. dnsdb lookups can now access SRV records.
1042 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1043 the configuration file.
1045 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1046 (-v) mode. This makes the output for a verbose queue run more intelligible.
1048 31. Added a use_postmaster feature to recipient callouts.
1050 32. Added the $body_zerocount variable, containing the number of binary zero
1051 bytes in the message body.
1053 33. The time of last modification of the "new" subdirectory is now used as the
1054 "mailbox time last read" when there is a quota error for a maildir
1057 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1059 35. Added +ignore_unknown as a special item in host lists.
1061 36. Code for decoding IPv6 addresses in host lists is now included, even if
1062 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1063 address was recognized as an IP address, but was then not correctly decoded
1064 into binary, causing unexpected and incorrect effects when compared with
1071 1. Very minor rewording of debugging text in manualroute to say "list of
1072 hosts" instead of "hostlist".
1074 2. If verify=header_syntax was set, and a header line with an unqualified
1075 address (no domain) and a large number of spaces between the end of the
1076 name and the colon was received, the reception process suffered a buffer
1077 overflow, and (when I tested it) crashed. This was caused by some obsolete
1078 code that should have been removed. The fix is to remove it!
1080 3. When running in the test harness, delay a bit after writing a bounce
1081 message to get a bit more predictability in the log output.
1083 4. Added a call to search_tidyup() just before forking a reception process. In
1084 theory, someone could use a lookup in the expansion of smtp_accept_max_
1085 per_host which, without the tidyup, could leave open a database connection.
1087 5. Added the variables $recipient_data and $sender_data which get set from a
1088 lookup success in an ACL "recipients" or "senders" condition, or a router
1089 "senders" option, similar to $domain_data and $local_part_data.
1091 6. Moved the writing of debug_print from before to after the "senders" test
1094 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1095 problems for message scanning, either using a data ACL, or using
1096 local_scan() because the Received: header was not generated till after they
1097 were called (in order to set the time as the time of reception completion).
1098 I have revised the way this works. The header is now generated after the
1099 body is received, but before the ACL or local_scan() are called. After they
1100 are run, the timestamp in the header is updated.
1106 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1107 before starting a queue runner without re-exec. This happened only when
1108 deliver_drop_privilege was set or when the Exim user was set to root. The
1109 effect of the bug was that timeouts during subsequent deliveries caused
1110 crashes instead of being properly handled. The handler is now left at its
1111 default (and expected) setting.
1113 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1114 message, again when deliver_drop_privilege is set or Exim is run as root.
1115 The bug described in (1) was not present in this case, but the tidying up
1116 of the other signals was missing. I have made the two cases consistent.
1118 3. The ignore_target_hosts setting on a manualroute router was being ignored
1119 for hosts that were looked up using the /MX notation.
1121 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1124 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1125 operated on the sender address. After changing the $sender_address to <>
1126 for the sender address verify, Exim was re-instated it as the original
1127 (before rewriting) address, but remembering that it had rewritten it, so it
1128 wasn't rewriting it again. This bug also had the effect of breaking the
1129 sender address verification caching when the sender address was rewritten.
1131 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1132 This has been changed so that if the ip literal address matches
1133 ignore_target_hosts, the router declines.
1135 7. Added expansion conditions match_domain, match_address, and match_local_
1136 part (NOT match_host).
1138 8. The placeholder for the Received: header didn't have a length field set.
1140 9. Added code to Exim itself and to exim_lock to test for a specific race
1141 condition that could lead to file corruption when using MBX delivery. The
1142 issue is with the lockfile that is created in /tmp. If this file is removed
1143 after a process has opened it but before that process has acquired a lock,
1144 there is the potential for a second process to recreate the file and also
1145 acquire a lock. This could lead to two Exim processes writing to the file
1146 at the same time. The added code performs the same test as UW imapd; it
1147 checks after acquiring the lock that its file descriptor still refers to
1148 the same named file.
1150 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1151 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1154 11. Added the smtp_active_hostname option. If used, this will typically be made
1155 to depend on the incoming interface address. Because $interface_address is
1156 not set up until the daemon has forked a reception process, error responses
1157 that can happen earlier (such as "too many connections") no longer contain
1160 12. If an expansion in a condition on a "warn" statement fails because a lookup
1161 defers, the "warn" statement is abandoned, and the next ACL statement is
1162 processed. Previously this caused the whole ACL to be aborted.
1164 13. Added the iplsearch lookup type.
1166 14. Added ident_timeout as a log selector.
1168 15. Added tls_certificate_verified as a log selector.
1170 16. Added a global option tls_require_ciphers (compare the smtp transport
1171 option of the same name). This controls incoming TLS connections.
1173 17. I finally figured out how to make tls_require_ciphers do a similar thing
1174 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1175 before starting the TLS session.
1177 18. Tabs are now shown as \t in -bP output.
1179 19. If the log selector return_path_on_delivery was set, Exim crashed when
1180 bouncing a message because it had too many Received: header lines.
1182 20. If two routers both had headers_remove settings, and the first one included
1183 a superfluous trailing colon, the final name in the first list and the
1184 first name in the second list were incorrectly joined into one item (with a
1185 colon in the middle).
1191 1. Added -C and -D options to the exinext utility, mainly to make it easier
1192 to include in the automated testing, but these could be helpful when
1193 multiple configurations are in use.
1195 2. The exinext utility was not formatting the output nicely when there was
1196 an alternate port involved in the retry record key, nor when there was a
1197 message id as well (for retries that were specific to a specific message
1198 and a specific host). It was also confused by IPv6 addresses, because of
1199 the additional colons they contain. I have fixed the IPv4 problem, and
1200 patched it up to do a reasonable job for IPv6.
1202 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1203 delivery, the log line now contains "pipelined" if PIPELINING was used.
1205 4. An SMTP transport process used to panic and die if the bind() call to set
1206 an explicit outgoing interface failed. This has been changed; it is now
1207 treated in the same way as a connect() failure.
1209 5. A reference to $sender_host_name in the part of a conditional expansion
1210 that was being skipped was still causing a DNS lookup. This no longer
1213 6. The def: expansion condition was not recognizing references to header lines
1214 that used bh_ and bheader_.
1216 7. Added the _cache feature to named lists.
1218 8. The code for checking quota_filecount in the appendfile transport was
1219 allowing one more file than it should have been.
1221 9. For compatibility with Sendmail, the command line option
1229 and sets the incoming protocol and host name (for trusted callers). The
1230 host name and its colon can be omitted when only the protocol is to be set.
1231 Note the Exim already has two private options, -pd and -ps, that refer to
1232 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1233 "s", but I don't think that's a major issue.
1235 10. A number of refactoring changes to the code, none of which should affect
1238 (a) The number of logging options was getting close to filling up the
1239 32-bit word that was used as a bit map. I have split them into two classes:
1240 those that are passed in the argument to log_write(), and those that are
1241 only ever tested independently outside of that function. These are now in
1242 separate 32-bit words, so there is plenty of room for expansion again.
1243 There is no change in the user interface or the logging behaviour.
1245 (b) When building, for example, log lines, the code previously used a
1246 macro that called string_cat() twice, in order to add two strings. This is
1247 not really sufficiently general. Furthermore, there was one instance where
1248 it was actually wrong because one of the argument was used twice, and in
1249 one call a function was used. (As it happened, calling the function twice
1250 did not affect the overall behaviour.) The macro has been replaced by a
1251 function that can join an arbitrary number of extra strings onto a growing
1254 (c) The code for expansion conditions now uses a table and a binary chop
1255 instead of a serial search (which was left over from when there were very
1256 few conditions). Also, it now recognizes conditions like "pam" even when
1257 the relevant support is not compiled in: a suitably worded error message is
1258 given if an attempt is made to use such a condition.
1260 11. Added ${time_interval:xxxxx}.
1262 12. A bug was causing one of the ddress fields not to be passed back correctly
1263 from remote delivery subprocesses. The field in question was not being
1264 subsequently used, so this caused to problems in practice.
1266 13. Added new log selectors queue_time and deliver_time.
1268 14. Might have fixed a bug in maildirsizefile handling that threw up
1269 "unexpected character" debug warnings, and recalculated the data
1270 unnecessarily. In any case, I expanded the warning message to give more
1273 15. Added the message "Restricted characters in address" to the statements in
1274 the default ACL that block characters like @ and % in local parts.
1276 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1277 Three changes have been made:
1279 (a) There was a serious bug; a negative response to MAIL caused the whole
1280 recipient domain to be cached as invalid, thereby blocking all messages
1281 to all local parts at the same domain, from all senders. This bug has
1282 been fixed. The domain is no longer cached after a negative response to
1283 MAIL if the sender used is not empty.
1285 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1288 (c) A new callout option, "use_sender" has been added for people who want
1289 the modified behaviour.
1295 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1298 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1299 indeed breaks things for older releases.
1301 3. Added additional logging to the case where there is a problem reading data
1302 from a filter that is running in a subprocess using a pipe, in order to
1303 try to track down a specific problem.
1305 4. Testing facility fudge: when running in the test harness and attempting
1306 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1307 getting "No route to host". Convert this to a timeout.
1309 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1312 6. Some OS don't have socklen_t but use size_t instead. This affects the
1313 fifth argument of getsockopt() amongst other things. This is now
1314 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1315 can be set for individual OS. I have set it for SunOS5, OSF1, and
1316 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1317 some earlier ones do not.
1319 7. Change 4.30/15 was not doing the test caselessly.
1321 8. The standard form for an IPv6 address literal was being rejected by address
1322 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1323 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1324 this, as well as the form without the "IPv6" on the front (but only when
1325 address literals are enabled, of course).
1327 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1329 10. Exim crashed if a message with an empty sender address specified by -f
1330 encountered a router with an errors_to setting. This could be provoked only
1331 by a command such as
1335 where an empty string was supplied; "<>" did not hit this bug.
1337 11. Installed PCRE release 4.5.
1339 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1340 remained set. It is now erased.
1342 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1343 times from message ids (which are base 36 rather than the normal 62).
1345 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1346 were being counted as actual protocol errors, and logged if the log
1347 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1348 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1349 and DATA following a set of rejected RCPTs do not count as protocol errors.
1350 In other words, Exim assumes they were pipelined, though this may not
1351 actually be the case. Of course, in all cases the client gets an
1352 appropriate error code.
1354 15. If a lookup fails in an ACL condition, a message about the failure may
1355 be available; it is used if testing the ACL cannot continue, because most
1356 such messages specify what the cause of the deferral is. However, some
1357 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1358 that caused an old message to be retained and used if a later statement
1359 caused a defer, replacing the real cause of the deferral.
1361 16. If an IP address had so many PTR records that the DNS lookup buffer
1362 was not large enough to hold them, Exim could crash while trying to process
1363 the truncated data. It now detects and logs this case.
1365 17. Further to 4.21/58, another change has been made: if (and only if) the
1366 first line of a message (the first header line) ends with CRLF, a bare LF
1367 in a subsequent header line has a space inserted after it, so as not to
1368 terminate the header.
1370 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1371 unnecessarily, used atoi() instead of strtol(), and didn't check the
1372 termination when getting file sizes from file names by regex.
1374 19. Completely re-implemented the support for maildirsize files, in the light
1375 of a number of problems with the previous contributed implementation
1376 (4.30/29). In particular:
1378 . If the quota is zero, the maildirsize file is maintained, but no quota is
1381 . If the maildir directory does not exist, it is created before any attempt
1382 to write a maildirsize file.
1384 . The quota value in the file is just a cache; if the quota is changed in
1385 the transport, the new value overrides.
1387 . A regular expression is available for excluding directories from the
1390 20. The autoreply transport checks the characters in options that define the
1391 message's headers; it allows continued headers, but it was checking with
1392 isspace() after an embedded newline instead of explicitly looking for a
1395 21. If all the "regular" hosts to which an address was routed had passed their
1396 expiry times, and had not reached their retry times, the address was
1397 bounced, even if fallback hosts were defined. Now Exim should go on to try
1400 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1401 equivalent code in the SMTP transport. Some hosts send humungous responses
1402 to HELO/EHLO, more than 1024 it seems.
1404 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1405 gives compiler warnings in some environments. I've now done it "properly",
1408 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1409 (because of problems with the built-in one) was declared to return uschar *
1410 instead of char *, causing compiler failure.
1412 25. Fixed a file descriptor leak when processing alias/forward files.
1414 26. Fixed a minor format string issue in dbfn.c.
1416 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1418 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1419 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1422 29. When a sender address is verified, it is cached, to save repeating the test
1423 when there is more than one recipient in a message. However, when the
1424 verification involves a callout, it is possible for different callout
1425 options to be set for different recipients. It is too complicated to keep
1426 track of this in the cache, so now Exim always runs a verification when a
1427 callout is required, relying on the callout cache for the optimization.
1428 The overhead is duplication of the address routing, but this should not be
1431 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1432 to be verified with callout=postmaster, and the main callout worked but the
1433 postmaster check failed, the verification correctly failed. However, if a
1434 subsequent RCPT command asked for sender verification *without* the
1435 postmaster check, incorrect caching caused this verification also to fail,
1438 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1439 it was not caching the DNS options (qualify_single, search_parents) that
1440 were used when the lookup failed. A subsequent lookup with different
1441 options therefore always gave the same answer, though there were cases
1442 where it should not have. (Example: a "domains = !$mx_any" option on a
1443 dnslookup router: the "domains" option is always processed without any
1444 widening, but the router might have qualify_single set.) Now Exim uses the
1445 cached value only when the same options are set.
1447 32. Added John Jetmore's "exipick" utility to the distribution.
1449 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1450 than a timeout (e.g. a certificate is required, and is not provided), an
1451 Exim server now closes the connection immediately. Previously it waited for
1452 the client to close - but if the client is SSL, it seems that they each
1453 wait for each other, leading to a delay before one of them times out.
1455 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1456 maintained 0.8.x compatibility because I don't think many are using it, and
1457 it is clearly obsolete.
1459 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1462 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1463 client certificate was expired. A simple patch fixes this, though I don't
1464 understand the full logic of why the verify callback is called multiple
1467 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1468 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1469 which causes problems with some clients (such as the Certicom SSL Plus
1470 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1471 disables the coutermeasure allowing Eudora to connect."
1473 38. Exim was not checking that a write() to a log file succeeded. This could
1474 lead to Bad Things if a log got too big, in particular if it hit a file
1475 size limit. Exim now panics and dies if it cannot write to a log file, just
1476 as it does if it cannot open a log file.
1478 39. Modified OS/Makefile-Linux so that it now contains
1480 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1482 The two -D definitions ensure that Exim is compiled with large file
1483 support, which makes it possible to handle log files that are bigger than
1486 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1487 instance) a domain was checked against a named list that involved a lookup,
1488 causing $domain_data to be set, then another domain was checked against the
1489 same list, then the first domain was re-checked, the value of $domain_data
1490 after the final check could be wrong. In particular, if the second check
1491 failed, it could be set empty. This bug probably also applied to
1494 41. The strip_trailing_dot option was not being applied to the address given
1495 with the -f command-line option.
1497 42. The code for reading a message's header from the spool was incrementing
1498 $received_count, but never initializing it. This meant that the value was
1499 incorrect (doubled) while delivering a message in the same process in which
1500 it was received. In the most common configuration of Exim, this never
1501 happens - a fresh exec is done - but it can happen when
1502 deliver_drop_privilege is set.
1504 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1505 it now includes up to 150 characters of the unexpected data in the log
1508 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1509 and building data strings. The size of both of these buffers was 10 000
1510 bytes - far larger than anybody would *ever* want, thought I. Needless to
1511 say, somebody hit the limit. I have increased the maximum line length to
1512 20 000 and the maximum data length of concatenated lines to 100 000. I have
1513 also fixed two bugs, because there was no checking on these buffers. Tsk,
1514 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1515 buffer is too small.
1517 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1518 lsearch lookups. Now it does.
1520 46. When parsing a route_list item in a manualroute router, a fixed-length
1521 buffer was used for the list of hosts. I made this 1024 bytes long,
1522 thinking that nobody would ever have a list of hosts that long. Wrong.
1523 Somebody had a whole pile of complicated expansion conditions, and the
1524 string was silently truncated, leading to an expansion error. It turns out
1525 that it is easier to change to an unlimited length (owing to other changes
1526 that have happened since this code was originally written) than to build
1527 structure for giving a limitation error. The length of the item that
1528 expands into the list of hosts is now unlimited.
1530 47. The lsearch lookup could not handle data where the length of text line was
1531 more than 4095 characters. Such lines were truncated, leading to shortened
1532 data being returned. It should now handle lines of any length.
1534 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1535 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1538 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1539 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1540 no longer happens in comment lines. A list of replacements is now placed
1541 at the head of all of the source files, except those whose only change is
1542 to replace PERL_COMMAND in the very first #! line.
1544 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1545 messages on the queue, with a bottom-up merge sort, using code contributed
1546 by Michael Haardt. This should make operations like -bp somewhat faster on
1547 large queues. It won't affect queue runners, except when queue_run_in_order
1550 51. Installed eximstats 1.31 in the distribution.
1552 52. Added support for SRV lookups to the dnslookup router.
1554 53. If an ACL referred to $message_body or $message_body_end, the value was not
1555 reset for any messages that followed in the same SMTP session.
1557 54. The store-handling optimization for building very long strings was not
1558 differentiating between the different store pools. I don't think this
1559 actually made any difference in practice, but I've tidied it.
1561 55. While running the routers to verify a sender address, $sender_address
1562 was still set to the sender address. This is wrong, because when routing to
1563 send a bounce to the sender, it would be empty. Therefore, I have changed
1564 it so that, while verifying a sender address, $sender_address is set to <>.
1565 (There is no change to what happens when verifying a recipient address.)
1567 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1568 target A or AAAA records (if not already returned) without resetting the
1569 qualify_single or search_parents options of the DNS resolver. These are
1570 inappropriate in this case because the targets of MX and SRV records must
1571 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1572 target that, when qualified, matched something in the local domain. These
1573 two options are now turned off when doing these lookups.
1575 57. It seems that at least some releases of Reiserfs (which does not have the
1576 concept of a fixed number of inodes) returns zero and not -1 for the
1577 number of available inodes. This interacted badly with check_spool_inodes,
1578 which assumed that -1 was the "no such thing" setting. What I have done is
1579 to check that the total number of inodes is greater than zero before doing
1580 the test of how many are available.
1582 58. When a "warn" ACL statement has a log_message modifier, the message is
1583 remembered, and not repeated. This is to avoid a lot of repetition when a
1584 message has many recipients that cause the same warning to be written.
1585 Howewer, Exim was preserving the list of already written lines for an
1586 entire SMTP session, which doesn't seem right. The memory is now reset if a
1587 new message is started.
1589 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1590 debugging output unless log_rewrite was also set.
1592 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1593 of (int)(handle) when we know that handle contains (void *)(-1).
1595 61. The Exim daemon panic-logs an error return when it closes the incoming
1596 connection. However "connection reset by peer" seems to be common, and
1597 isn't really an error worthy of noting specially, so that particular error
1600 62. When Exim is trying to find all the local interfaces, it used to panic and
1601 die if the ioctl to get the interface flags failed. However, it seems that
1602 on at least one OS (Solaris 9) it is possible to have an interface that is
1603 included in the list of interfaces, but for which you get a failure error
1604 for this call. This happens when the interface is not "plumbed" into a
1605 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1606 failure of the "get flags" call assumes that the interface is down.
1608 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1609 makes life easier for people who are doing arithmetic on fields extracted
1610 from dates, where you often get leading zeros that should not be
1611 interpreted as octal.
1613 64. Added qualify_domain to the redirect router, to override the global
1616 65. If a pathologically long header line contained very many addresses (the
1617 report of this problem mentioned 10 000) and each of them was rewritten,
1618 Exim could use up a very large amount of memory. (It kept on making new
1619 copies of the header line as it rewrote, and never released the old ones.)
1620 At the expense of a bit more processing, the header rewriting function has
1621 been changed so that it no longer eats memory in this way.
1623 66. The generation of the Received: header has been moved from the time that a
1624 message starts to be received, to the time that it finishes. The timestamp
1625 in the Received: header should now be very close to that of the <= log
1626 line. There are two side-effects of this change:
1628 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1629 logged header lines no longer include the local Received: line, because
1630 it has not yet been created. The same applies to a copy of the message
1631 that is returned to a non-SMTP sender when a message is rejected.
1633 (b) When a filter file is tested using -bf, no additional Received: header
1634 is added to the test message. After some thought, I decided that this
1637 This change does not affect the value of $received_for. It is still set
1638 after address rewriting, but before local_scan() is called.
1640 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1642 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1643 gave an unhelpful panic error message, and a defer error. I have managed to
1644 change this behaviour so that it now rejects any supplied certificate,
1645 which seems right, as the list of acceptable certificates is empty.
1647 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1648 gave an unhelpful defer error. I have not managed to make this reject any
1649 supplied certificates, but the error message it gives is "no certificate
1650 supplied", which is not helpful.
1652 70. exigrep's output now also includes lines that are not associated with any
1653 message, but which match the given pattern. Implemented by a patch from
1654 Martin Sluka, which also tidied up the Perl a bit.
1656 71. Recipient callout verification, like sender verification, was using <> in
1657 the MAIL FROM command. This isn't really the right thing, since the actual
1658 sender may affect whether the remote host accepts the recipient or not. I
1659 have changed it to use the actual sender in the callout; this means that
1660 the cache record is now keyed on a recipient/sender pair, not just the
1661 recipient address. There doesn't seem to be a real danger of callout loops,
1662 since a callout by the remote host to check the sender would use <>.
1663 [SEE ABOVE: changed after hitting problems.]
1665 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1666 temporary errors. However, in the case of such a code being given after
1667 the end of a data transmission (i.e. after ".") Exim was failing to write
1668 a retry record for the message. (Yes, there was some broken host that was
1669 actually sending 8xx at this point.)
1671 73. An unknown lookup type in a host list could cause Exim to panic-die when
1672 the list was checked. (An example that provoked this was putting <; in the
1673 middle of a list instead of at the start.) If this happened during a DATA
1674 ACL check, a -D file could be left lying around. This kind of configuration
1675 error no longer causes Exim to die; instead it causes a defer errror. The
1676 incident is still logged to the main and panic logs.
1678 74. Buglet left over from Exim 3 conversion. The message "too many messages
1679 in one connection" was written to the rejectlog but not the mainlog, except
1680 when address rewriting (yes!) was being logged.
1682 75. Added write_rejectlog option.
1684 76. When a system filter was run not as root (that is, when system_filter_user
1685 was set), the values of the $n variables were not being returned to the
1686 main process; thus, they were not subsequently available in the $sn
1689 77. Added +return_path_on_delivery log selector.
1691 78. A connection timeout was being treated differently from recipients deferred
1692 when testing hosts_max_try with a message that was older than the host's
1693 retry timeout. (The host should not be counted, thus allowing all hosts to
1694 be tried at least once before bouncing.) This may have been the cause of an
1695 occasionally reported bug whereby a message would remain on the queue
1696 longer than the retry timeout, but would be bounced if a delivery was
1697 forced. I say "may" because I never totally pinned down the problem;
1698 setting up timeout/retry tests is difficult. See also the next item.
1700 79. The ultimate address timeout was not being applied to errors that involved
1701 a combination of host plus message (for example, a timeout on a MAIL
1702 command). When an address resolved to a number of possible hosts, and they
1703 were not all tried for each delivery (e.g. because of hosts_max_try), a
1704 message could remain on the queue longer than the retry timeout.
1706 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1709 81. Fixed an obscure SMTP outgoing bug which required at least the following
1710 conditions: (a) there was another message waiting for the same server;
1711 (b) the server returned 5xx to all RCPT commands in the first message so
1712 that the message was not completed; (c) the server dropped the connection
1713 or gave a negative response to the RSET that Exim sends to abort the
1714 transaction. The observed case was a dropped connection after DATA that had
1715 been sent in pipelining mode. That is, the server had advertised PIPELINING
1716 but was not implementing it correctly. The effect of the bug was incorrect
1717 behaviour, such as trying another host, and this could lead to a crash.
1723 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1724 and daemon.c were passed as pointers to ints; they should have been
1725 pointers to socklen_t variables (which are typically unsigned ints).
1727 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1730 3. Fixed a really odd bug that affected only the testing scheme; patching a
1731 certain fixed string in the binary changed the value of another string that
1732 happened to be identical to the end of the original first string.
1734 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1735 name", it returns that address as the IP address. On some operating
1736 systems (e.g. Solaris), it also passes back the IP address string as the
1737 "host name". However, on others (e.g. Linux), it passes back an empty
1738 string. Exim wasn't checking for this, and was changing the host name to an
1739 empty string, assuming it had been canonicized.
1741 5. Although rare, it is permitted to have more than one PTR record for a given
1742 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1743 all the names associated with an address, because they do in Solaris.
1744 However, it seems that they do not in Linux for data that comes from the
1745 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1746 I found this out when I moved to a new Linux workstation and tried to run
1747 the Exim test suite.
1749 To get round this problem I have changed the code so that it now does its
1750 own call to the DNS to look up PTR records when searching for a host name.
1751 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1752 addresses that are only in /etc/hosts are still found.
1754 This behaviour is, however, controlled by an option called host_lookup_
1755 order, which defaults to "bydns:byaddr". If people want to use the other
1756 order, or indeed, just use one or the other means of lookup, they can
1757 specify it in this variable.
1759 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1760 some operating systems, this comes back from gethostbyaddr() as an empty
1761 string, and this is what Exim used to test for. However, it seems that in
1762 other systems, "." is yielded. Exim now tests for this case too.
1764 7. The values of check_spool_space and check_log_space are now held internally
1765 as a number of kilobytes instead of an absolute number of bytes. If a
1766 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1767 kilobyte. This means that much larger values can be stored.
1769 8. Exim monitor: an attempt to get the action menu when not actually pointing
1770 at a message produces an empty menu entitled "No message selected". This
1771 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1772 no entries in it ("Shell widget menu has zero width and/or height"). So I
1773 have added a single, blank menu entry in this case.
1775 9. Added ${quote_local_part.
1777 10. MIME decoding is now applied to the contents of Subject: header lines when
1780 11. Now that a reference to $sender_host_address automatically causes a reverse
1781 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1782 host lookup before query-style lookups in lists that might use this
1783 variable. This has therefore been abolished, and the "net-" prefix is no
1784 longer necessary for query-style lookups.
1786 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1787 have been a typo for LFLAGS, so it has been changed.
1789 13. The install script calls Exim with "-C /dev/null" in order to find the
1790 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1791 to be output. Howeve, since Exim outputs its version number before the
1792 error, it didn't break the script. It just looked ugly. I fixed this by
1793 always allowing "-C /dev/null" if the caller is root.
1795 14. Ignore overlarge ACL variable number when reading spool file - insurance
1796 against a later release with more variables having written the file.
1798 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1799 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1800 this, as well as the form without the "IPv6" on the front.
1802 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1803 OS/Makefile-Darwin file.
1805 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1806 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1808 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1809 variable. However, some OS use ints for inodes. I've added cast to long int
1810 to get rid of the compiler warning.
1812 19. I had forgotten to lock out "/../" in configuration file names when
1813 ALT_CONFIG_PREFIX was set.
1815 20. Routers used for verification do not need to specify transports. However,
1816 if such a router generated a host list, and callout was configured, Exim
1817 crashed, because it could not find a port number from the (non-existent)
1818 transport. It now assumes port 25 in this circumstance.
1820 21. Added the -t option to exigrep.
1822 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1823 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1824 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1825 them is set, LOOKUP_LSEARCH is forced.
1827 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1829 24. Added sender and host information to the "rejected by local_scan()" log
1830 line; previously there was no indication of these.
1832 25. Added .include_if_exists.
1834 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1835 Linux. It turns out that not all file systems support this. Apparently some
1836 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1837 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1838 not supported on the file descriptor, is now ignored when Exim is trying to
1839 sync a directory. This applies only to Linux.
1841 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1843 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1844 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1847 29. Added support for maildirsize files from supplied patch (modified a bit).
1849 30. The use of :fail: followed by an empty string could lead Exim to respond to
1850 sender verification failures with (e.g.):
1852 550 Verification failed for <xxx>
1853 550 Sender verify failed
1855 where the first response line was missing the '-' that indicates it is not
1856 the final line of the response.
1858 31. The loop for finding the name of the user that called Exim had a hardwired
1859 limit of 10; it now uses the value of finduser_retries, which is used for
1860 all other user lookups.
1862 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1865 33. Exim was neglecting to zero errno before one call of strtol() when
1866 expanding a string and expecting an integer value. On some systems this
1867 resulted in spurious "integer overflow" errors. Also, it was casting the
1868 result into an int without checking.
1870 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1871 did not work. The code looks as if it has *never* worked, though it appears
1872 to have been documented since at least releast 1.62. I have made it work.
1874 35. The "timeout_DNS" error in retry rules, also documented since at least
1875 1.62, also never worked. As it isn't clear exactly what this means, and
1876 clearly it isn't a major issue, I have abolished the feature by treating it
1877 as "timeout", and writing a warning to the main and panic logs.
1879 36. The display of retry rules for -brt wasn't always showing the error code
1882 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1883 timeout_connect_A, timeout_connect_MX.
1885 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1886 to the empty sender.
1888 39. The daemon was not analysing the content of -oX till after it had closed
1889 stderr and disconnected from the controlling terminal. This meant that any
1890 syntax errors were only noted on the panic log, and the return code from
1891 the command was 0. By re-arranging the code a little, I've made the
1892 decoding happen first, so such errors now appear on stderr, and the return
1893 code is 1. However, the actual setting up of the sockets still happens in
1894 the disconnected process, so errors there are still only recorded on the
1897 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1898 connections (as happens on some IP stacks) was logged at start up time as
1899 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1900 it from "IPv6 and IPv4", which means that two separate sockets are being
1903 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1904 says whether AF_INET or AF_INET6 was passed as an argument.
1906 42. Exiwhat output was messed up when time zones were included in log
1909 43. Exiwhat now gives more information about the daemon's listening ports,
1910 and whether -tls-on-connect was used.
1912 44. The "port" option of the smtp transport is now expanded.
1914 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1915 silently ignored. Now an error message is written to the main and panic
1918 46. There's a new ACL modifier called "logwrite" which writes to a log file
1919 as soon as it is encountered.
1921 47. Added $local_user_uid and $local_user_gid at routing time.
1923 48. Exim crashed when trying to verify a sender address that was being
1926 49. Exim was recognizing only a space character after ".include". It now also
1927 recognizes a tab character.
1929 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1930 extracting the relevant information from the specification. The man page no
1931 longer contains scrambled data for the -d option, and I've added a section
1932 at the front about calling Exim under different names.
1934 51. Added "extra_headers" argument to the "mail" command in filter files.
1936 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1939 53. Installed eximstats 1.29.
1941 54. Added transport_filter_timeout as a generic transport option.
1943 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1944 Cc: header lines. This was required by RFC 822, but it not required by RFC
1947 56. Exim used to add From:, Date:, and Message-Id: header lines to any
1948 incoming messages that did not have them. Now it does so only if the
1949 message originates locally, that is, if there is no associated remote host
1950 address. When Resent- header lines are present, this applies to the Resent-
1951 lines rather than the non-Resent- lines.
1953 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
1954 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
1956 58. Messages for configuration errors now include the name of the main
1957 configuration file - useful now that there may be more than one file in a
1958 list (.included file names were always shown).
1960 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1961 for those rare installations that do not start the daemon as root or run it
1962 setuid root. I've cut out the call to initgroups() if the daemon is not
1965 60. The Exim user and group can now be bound into the binary as text strings
1966 that are looked up at the start of Exim's processing.
1968 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
1970 62. Added $mailstore_basename variable.
1972 63. Installed patch to sieve.c from Michael Haardt.
1974 64. When Exim failed to open the panic log after failing to open the main log,
1975 the original message it was trying to log was written to stderr and debug
1976 output, but if they were not available (the usual case in production), it
1977 was lost. Now it is written to syslog before the two lines that record the
1978 failures to open the logs.
1980 65. Users' Exim filters run in subprocesses under the user's uid. It is
1981 possible for a "deliver" command or an alias in a "personal" command to
1982 provoke an address rewrite. If logging of address rewriting is configured,
1983 this fails because the process is not running as root or exim. There may be
1984 a better way of dealing with this, but for the moment (because 4.30 needs
1985 to be released), I have disabled address rewrite logging when running a
1986 filter in a non-root, non-exim process.
1992 1. The buildconfig auxiliary program wasn't quoting the value set for
1993 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
1994 not defined. This bug was masked in 4.22 by the effect that was fixed in
1997 2. Some messages that were rejected after a message id was allocated were
1998 shown as "incomplete" by exigrep. It no longer does this for messages that
1999 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2001 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2002 have allow_domain_literals set, the ID did not get logged in the <= line.
2003 Domain literals are now always recognized in Message-ID: header lines.
2005 4. The first argument for a ${extract expansion item is the key name or field
2006 number. Leading and trailing spaces in this item were not being ignored,
2007 causing some misleading effects.
2009 5. When deliver_drop_privilege was set, single queue runner processes started
2010 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2011 same command in the process it spins off) were not dropping privilege.
2013 6. When the daemon running as "exim" started a queue runner, it always
2014 re-executed Exim in the spun-off process. This is a waste of effort when
2015 deliver_drop_privilege is set. The new process now just calls the
2016 queue-runner function directly.
2022 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2025 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2026 ignored. Though the use of -oP was forcing the writing of a pid file, it
2027 was always written to the default place.
2029 3. If the message "no IP address found for host xxxx" is generated during
2030 incoming verification, it is now followed by identification of the incoming
2031 connection (so you can more easily find what provoked it).
2033 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2035 5. Added some features to "harden" Exim a bit more against certain attacks:
2037 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2038 be put in Local/Makefile. This is like the never_users runtime option,
2039 but it cannot be overridden. The default setting is "root".
2041 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2042 prefix string with which any file named in a -C command line option
2045 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2046 is retained for -C and -D only if the caller of Exim is root. Without
2047 it, the exim user may also use -C and -D and retain privilege.
2049 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2050 command line option is disabled.
2052 6. Macro names set by the -D option must start with an upper case letter, just
2053 like macro names defined in the configuration file.
2055 7. Added "dereference=" facility to LDAP.
2057 8. Two instances of the typo "uknown" in the source files are fixed.
2059 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2060 the Configure-Makefile script screwed up while processing it.
2062 10. Incorporated PCRE 4.4.
2064 11. The SMTP synchronization check was not operating right at the start of an
2065 SMTP session. For example, it could not catch a HELO sent before the client
2066 waited for the greeting. There is now a check for outstanding input at the
2067 point when the greeting is written. Because of the duplex, asynchronous
2068 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2069 way, but not yet received, when the check is performed.
2071 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2072 on TCP/IP sockets, because this apparently causes some broken clients to
2075 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2076 unchanged) from the Cygwin maintainer.
2078 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2079 was supported instead of testing for mbx. Effectively a typo.
2081 15. The spa authenticator server code was not checking that the input it
2082 received was valid base64.
2084 16. The debug output line for the "set" modifier in ACLs was not showing the
2085 name of the variable that was being set.
2087 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2089 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2090 lookup on its own. Something else was needed to trigger the lookup. For
2091 example, a match in host_lookup or the need for a host name in a host list.
2092 Now, if $sender_host_name is referenced and the host name has not yet been
2093 looked up, a lookup is performed. If the lookup fails, the variable remains
2094 empty, and $host_lookup_failed is set to "1".
2096 19. Added "eqi" as a case-independent comparison operator.
2098 20. The saslauthd authentication condition could segfault if neither service
2099 nor realm was specified.
2101 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2102 error message that was logged was misleading, and incoming SMTP
2103 connections were dropped. The message is now more accurate, and temporary
2104 errors are given to SMTP connections.
2106 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2107 (except RSET) with a 421 error, until QUIT is received. However, it was
2108 failing to send a response to QUIT.
2110 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2111 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2112 ACL is now run after the helo_try_verify_hosts code.
2114 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2115 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2116 case-independent, so other case variants are also recognized). Apparently
2117 some systems use these upper case variants.
2119 25. If more than two messages were waiting for the same host, and a transport
2120 filter was specified for the transport, Exim sent two messages over the
2121 same TCP/IP connection, and then failed with "socket operation on non-
2122 socket" when it tried to send the third.
2124 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2126 27. The extern definition of crypt16() in expand.c was not being excluded when
2127 the OS had its own crypt16() function.
2129 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2130 as a preferred synonym for return_size_limit, both as an option and as an
2133 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2135 30. Changed the default configuration ACL to relax the local part checking rule
2136 for addresses that are not in any local domains. For these addresses,
2137 slashes and pipe symbols are allowed within local parts, but the sequence
2138 /../ is explicitly forbidden.
2140 31. SPA server authentication was not clearing the challenge buffer before
2143 32. log_message in a "warn" ACL statement was writing to the reject log as
2144 well as to the main log, which contradicts the documentation and doesn't
2145 seem right (because no rejection is happening). So I have stopped it.
2147 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2148 However, I am unable to do any testing of this.
2150 34. Fixed an infelicity in the appendfile transport. When checking directories
2151 for a mailbox, to see if any needed to be created, it was accidentally
2152 using path names with one or more superfluous leading slashes; tracing
2153 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2155 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2156 contents are added to the log line that is written for every discarded
2157 recipient. (Previously a log_message setting was ignored.)
2159 36. The ${quote: operator now quotes the string if it is empty.
2161 37. The install script runs exim in order to find its version number. If for
2162 some reason other than non-existence or emptiness, which it checks, it
2163 could not run './exim', it was installing it with an empty version number,
2164 i.e. as "exim-". This error state is now caught, and the installation is
2167 38. An argument was missing from the function that creates an error message
2168 when Exim fails to connect to the socket for saslauthd authentication.
2169 This could cause Exim to crash, or give a corrupted message.
2171 39. Added isip, isip4, and isip6 to ${if conditions.
2173 40. The ACL variables $acl_xx are now saved with the message, and can be
2174 accessed later in routers, transports, and filters.
2176 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2177 strings in the file are not string-expanded.
2179 42. If a MAIL command specified a SIZE value that was too large to fit into an
2180 int variable, the check against message_size_limit failed. Such values are
2181 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2182 day this will have to be increased, but I don't think I want to be around
2183 when emails are that large.
2190 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2191 iconv() is not standard in FreeBSD.
2193 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2194 IPv6 enabled. The observed symptom was a segmentation fault on return from
2195 the function os_common_find_running_interfaces() in src/os.c.
2197 3. In the check_special_case() function in daemon.c I had used "errno" as an
2198 argument name, which causes warnings on some systems. This was basically a
2199 typo, since it was named "eno" in the comments!
2201 4. The code that waits for the clock to tick (at a resolution of some fraction
2202 of a second) so as to ensure message-id uniqueness was always waiting for
2203 at least one whole tick, when it could have waited for less. [This is
2204 almost certainly not relevant at current processor speeds, where it is
2205 unlikely to ever wait at all. But we try to future-proof.]
2207 5. The function that sleeps for a time interval that includes fractions of a
2208 second contained a race. It did not block SIGALRM between setting the
2209 timer, and suspending (a couple of lines later). If the interval was short
2210 and the sigsuspend() was delayed until after it had expired, the suspension
2211 never ended. On busy systems this could lead to processes getting stuck for
2214 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2215 process, before it forks any delivery processes. The open lookup caching
2216 mechanism meant that the open file or database connection was passed into
2217 the delivery process. The problem was that delivery processes always tidy
2218 up cached lookup data. This could cause a problem for the next delivery
2219 process started by the queue runner, because the external queue runner
2220 process does not know about the closure. So the next delivery process
2221 still has data in the lookup cache. In the case of a file lookup, there was
2222 no problem because closing a file descriptor in a subprocess doesn't affect
2223 the parent. However, if the lookup was caching a connection to a database,
2224 the connection was closed, and the second delivery process was likely to
2225 see errors such as "PGSQL: query failed: server closed the connection
2226 unexpectedly". The problem has been fixed by closing all cached lookups
2227 in a queue runner before running a delivery process.
2229 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2230 seem to have the "const" qualifier which it has on other OS. I've
2233 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2234 *of the same type* (client or server) with the same public name that an
2235 error should be diagnosed.
2237 9. When Exim looked up a host name for an IP address, but failed to find the
2238 original IP address when looking up the host name (a safety check), it
2239 output the message "<ip address> does not match any IP for NULL", which was
2240 confusing, to say the least. The bug was that the host name should have
2241 appeared instead of "NULL".
2243 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2244 user that is built into the binary, and the -C or -D options is used, root
2245 privilege is dropped before the configuration file is read. In addition,
2246 logging is switched to stderr instead of the normal log files. If the
2247 configuration then re-defines the Exim user, the unprivileged environment
2248 is probably not what is expected, so Exim logs a panic warning message (but
2251 However, if deliver_drop_privilege is set, the unprivileged state may well
2252 be exactly what is intended, so the warning has been cut out in that case,
2253 and Exim is allowed to try to write to its normal log files.
2259 1. smtp_return_error_details was not giving details for temporary sender
2260 or receiver verification errors.
2262 2. Diagnose a configuration error if two authenticators have the same public
2265 3. Exim used not to create the message log file for a message until the first
2266 delivery attempt. This could be confusing when incoming messages were held
2267 for policy or load reasons. The message log file is now created at the time
2268 the message is received, and an initial "Received" line is written to it.
2270 4. The automatically generated man page for command line options had a minor
2271 bug that caused no ill effects; however, a more serious problem was that
2272 the procedure for building the man page automatically didn't always
2273 operate. Consequently, release 4.20 contains an out-of-date version. This
2274 shouldn't happen again.
2276 5. When building Exim with embedded Perl support, the script that builds the
2277 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2278 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2280 6. The freeze_tell option was not being used for messages that were frozen on
2281 arrival, either by an ACL or by local_scan().
2283 7. Added the smtp_incomplete_transaction log selector.
2285 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2286 was accepting AUTH without a new EHLO.
2288 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2289 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2292 10. Logging of TCP/IP connections (when configured) now happens in the main
2293 daemon process instead of the child process, so that the TCP/IP connection
2294 count is more accurate (but it can never be perfect).
2296 11. The use of "drop" in a nested ACL was not being handled correctly in the
2297 outer ACL. Now, if condition failure induced by the nested "drop" causes
2298 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2299 or "require"), the connection is dropped.
2301 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2302 that yield "discard" can now be used with an "accept" or a "discard" verb,
2303 but an error is generated for any others (because I can't see a useful way
2304 to define what should happen).
2306 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2307 are now processed in the same way as lines in the Exim configuration file.
2308 In particular, continuation lines are supported.
2310 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2312 15. Added -ti meaning -t -i.
2314 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2315 domains, and warn by logging if others are found.
2317 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2318 returned from GIFCONF when Exim is trying to find the list of interfaces on
2319 a host. The code in os.c has been modified to copy each ifreq to an aligned
2320 structure in all cases.
2322 Also, in some cases, the returned ifreq's were being copied to a 'struct
2323 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2324 means the last couple of bytes of an IPv6 address could be chopped if the
2325 ifreq contained only a normal sockaddr (14 bytes storage).
2327 18. Named domain lists were not supported in the hosts_treat_as_local option.
2328 An entry such as +xxxx was not recognized, and was treated as a literal
2331 19. Ensure that header lines added by a DATA ACL are included in the reject log
2332 if the ACL subsequently rejects the message.
2334 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2335 MD5 (which is deprecated).
2337 21. When testing a filter file using -bf, Exim was writing a message when it
2338 took the sender from a "From " line in the message, but it was not doing so
2339 when it took $return_path from a Return-Path: header line. It now does.
2341 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2342 with a valid header line field name (a series of printing characters
2343 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2345 23. Changed "disc" in the source to "disk" to conform to the documentation and
2346 the book and for uniformity.
2348 24. Ignore Sendmail's -Ooption=value command line item.
2350 25. When execve() failed while trying to run a command in a pipe transport,
2351 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2352 could be confused with a return value of 69 from the command itself. This
2353 has been changed to 127, the value the shell returns if it is asked to run
2354 a non-existent command. The wording for the related log line suggests a
2355 non-existent command as the problem.
2357 26. If received_header_text expands to an empty string, do not add a Received:
2358 header line to the message. (Well, it adds a token one on the spool, but
2359 marks it "old" so that it doesn't get used or transmitted.)
2361 27. Installed eximstats 1.28 (addition of -nt option).
2363 28. There was no check for failure on the call to getsockname() in the daemon
2364 code. This can fail if there is a shortage of resources on the system, with
2365 ENOMEM, for example. A temporary error is now given on failure.
2367 29. Contrary to the C standard, it seems that in some environments, the
2368 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2369 program. Exim now does this explicitly; it affects the formatting of
2370 timestamps using strftime().
2372 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2373 complaints. I've now initialized all the variables, to avoid this.
2375 32. Header lines added by a system filter were not being "seen" during
2376 transport-time rewrites.
2378 33. The info_callback() function passed to OpenSSL is set up with type void
2379 (*)(SSL *, int, int), as described somewhere. However, when calling the
2380 function (actually a macro) that sets it up, the type void(*)() is
2381 expected. I've put in a cast to prevent warnings from picky compilers.
2383 34. If a DNS black list lookup found a CNAME record, but there were no A
2384 records associated with the domain it pointed at, Exim crashed.
2386 35. If a DNS black list lookup returned more than one A record, Exim ignored
2387 all but the first. It now scans all returned addresses if a particular IP
2388 value is being sought. In this situation, the contents of the
2389 $dnslist_value variable are a list of all the addresses, separated by a
2392 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2393 to accept a host name and all its aliases if the forward lookup for any of
2394 them yielded the IP address of the incoming connection. Now it accepts only
2395 those names whose forward lookup yields the correct IP address. Any other
2396 names are discarded. This closes a loophole whereby a rogue DNS
2397 administrator could create reverse DNS records to break through a
2398 wildcarded host restriction in an ACL.
2400 37. If a user filter or a system filter that ran in a subprocess used any of
2401 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2402 the wrong values were passed to the pipe command ($thisaddress had the
2403 value of $0, $0 had the value of $1, etc). This bug was introduced by
2404 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2406 38. Improved the line breaking for long SMTP error messages from ACLs.
2407 Previously, if there was no break point between 40 and 75 characters, Exim
2408 left the rest of the message alone. Two changes have been made: (a) I've
2409 reduced the minimum length to 35 characters; (b) if it can't find a break
2410 point between 35 and 75 characters, it looks ahead and uses the first one
2411 that it finds. This may give the occasional overlong line, but at least the
2412 remaining text gets split now.
2414 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2415 file descriptors might be low, and that setting 1000 would always raise it.
2416 It turns out that in some environments, the limit is already over 1000 and
2417 that lowering it causes trouble. So now Exim takes care not to decrease it.
2419 40. When delivering a message, the value of $return_path is set to $sender_
2420 address at the start of routing (routers may change the value). By an
2421 oversight, this default was not being set up when an address was tested by
2422 -bt or -bv, which affected the outcome if any router or filter referred to
2425 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2426 log only when "message" or "log_message" are set. However, if one of the
2427 conditions was an address verification, or a call to a nested ACL, the
2428 messages generated by the underlying test were being passed through. This
2429 no longer happens. The underlying message is available in $acl_verify_
2430 message for both "message" and "log_message" expansions, so it can be
2431 passed through if needed.
2433 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2434 new expansion $bh_ to give the encoded byte string without charset
2435 translation. Translation happens only if iconv() is available; HAVE_ICONV
2436 indicates this at build time. HEADERS_CHARSET gives the charset to
2437 translate to; headers_charset can change it in the configuration, and
2438 "headers charset" can change it in an individual filter file.
2440 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2441 that creates RFC 2047 encoded "words" labels them as that charset instead
2442 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2443 expansion operator; (ii) when Exim creates a From: line for a local
2444 message; (iii) when a header line is rewritten to include a "phrase" part.
2446 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2447 buggy, causing it to skip the first lines of messages whose message ID
2448 ended in 'D'. This would not have bitten before Exim release 4.14, because
2449 message IDs were unlikely to end in 'D' before then. The effect was to have
2450 incorrect size information for certain domains.
2452 45. #include "config.h" was missing at the start of the crypt16.c module. This
2453 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2456 46. If there was a timeout during a "random" callout check, Exim treated it as
2457 a failure of the random address, and carried on sending RSET and the real
2458 address. If the delay was just some slowness somewhere, the response to the
2459 original RCPT would be taken as a response to RSET and so on, causing
2460 mayhem of various kinds.
2462 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2463 when I implemented it. It didn't allow for the fact that some option values
2464 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2465 the right test for positive values.
2467 48. Domain names in DNS records are case-independent. Exim always looks them up
2468 in lower case. Some resolvers return domain names in exactly the case they
2469 appear in the zone file, that is, they may contain uppercase letters. Not
2470 all resolvers do this - some return always lower case. Exim was treating a
2471 change of case by a resolver as a change of domain, similar to a widening
2472 of a domain abbreviation. This triggered its re-routing code and so it was
2473 trying to route what was effectively the same domain again. This normally
2474 caused routing to fail (because the router wouldn't handle the domain
2475 twice). Now Exim checks for this case specially, and just changes the
2476 casing of the domain that it ultimately uses when it transmits the message
2479 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2482 50. If a filter generated a file delivery with a non-absolute name (possible if
2483 no home directory exists for the router), the forbid_file option was not
2486 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2487 the existing equality matching.
2489 52. Exim was using ints instead of ino_t variables in some places where it was
2490 dealing with inode numbers.
2492 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2493 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2494 called TMPDIR, and if it finds it is different, it changes its value.
2496 54. The smtp_printf() function is now made available to local_scan() so
2497 additional output lines can be written before returning. There is also an
2498 smtp_fflush() function to enable the detection of a dropped connection.
2499 The variables smtp_input and smtp_batched_input are exported to
2502 55. Changed the default runtime configuration: the message "Unknown user"
2503 has been removed from the ACL, and instead placed on the localuser router,
2504 using the cannot_route_message feature. This means that any verification
2505 failures that generate their own messages won't get overridden. Similarly,
2506 the "Unrouteable address" message that was in the ACL for unverifiable
2507 relay addresses has also been removed.
2509 56. Added hosts_avoid_esmtp to the smtp transport.
2511 57. The exicyclog script was not checking for the esoteric option
2512 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2513 will work only if exicyclog is run under the appropriate euid.
2515 58. Following a discussion on the list, the rules by which Exim recognises line
2516 endings on incoming messages have been changed. The -dropcr and drop_cr
2517 options are now no-ops, retained only for backwards compatibility. The
2518 following line terminators are recognized: LF CRLF CR. However, special
2519 processing applies to CR:
2521 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2522 nor a local message in the state where . is a terminator.
2524 (ii) If a bare CR is encountered in a header line, an extra space is added
2525 after the line terminator so as not to end the header. The reasoning
2526 behind this is that bare CRs in header lines are most likely either
2527 to be mistakes, or people trying to play silly games.
2529 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2530 was being incorrectly given as 18 bytes larger than it should have been.
2531 This is a VOB (very old bug).
2533 60. This may never have affected anything current, but just in case it has:
2534 When the local host is found other than at the start of a list of hosts,
2535 the local host, those with the same MX, and any that follow, are discarded.
2536 When the list in question was part of a longer list of hosts, the following
2537 hosts (not currently being processed) were also being discarded. This no
2538 longer happens. I'm not sure if this situation could ever has previously
2541 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2544 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2545 header. This is something that is recommended in a new Internet Draft, and
2546 is something that is documented as being done by Sendmail. There are two
2547 possible values. For messages generated by the autoreply transport, Exim
2550 Auto-Submitted: auto-replied
2552 whereas for all other generated messages (e.g. bounces) it adds
2554 Auto-Submitted: auto-generated
2556 63. The "personal" condition in filters now includes a test for the
2557 Auto-Submitted: header. If it contains the string "auto-" the message it
2558 not considered personal.
2560 64. Added rcpt_include_affixes as a generic transport option.
2562 65. Added queue_only_override (default true).
2564 66. Added the syslog_duplication option.
2566 67. If what should have been the first header line of a message consisted of
2567 a space followed by a colon, Exim was mis-interpreting it as a header line.
2568 It isn't of course - it is syntactically invalid and should therefore be
2569 treated as the start of the message body. The misbehaviour could have
2570 caused a number of strange effects, including loss of data in subsequent
2571 header lines, and spool format errors.
2573 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2574 client host had authenticated. This control can now be exercised by an ACL
2575 for more flexibility.
2577 69. By default, callouts do not happen when testing with -bh. There is now a
2578 variant, -bhc, which does actually run the callout code, including
2579 consulting and updating the callout cache.
2581 70. Added support for saslauthd authentication, courtesy of Alexander
2584 71. If statvfs() failed on the spool or log directories while checking their
2585 size for availability, Exim confusingly gave the error "space shortage".
2586 Furthermore, in debugging mode it crashed with a floating point exception.
2587 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2588 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2589 serious problem, Exim now writes to the main and panic logs when this
2590 happens, with details of the failure. It then refuses to accept the
2591 incoming message, giving the message "spool directory problem" or "log
2592 directory problem" with a 421 code for SMTP messages.
2594 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2595 0, 1, and 2 exist, because some OS complain for execs without them (see
2596 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2597 descriptors. However, the code omitted to check that the open succeeded,
2598 causing mysterious errors if for some reason the permissions on /dev/null
2599 got screwed. Now Exim writes a message to the main and panic logs, and
2600 bombs out if it can't open /dev/null.
2602 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2603 interact so that it is all more flexible. It is supposed to remain
2604 backwards compatible. Also added extra_local_interfaces.
2606 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2607 to bomb out with an assertion failure - to the client this appears as a
2608 connection drop. This problem occurs in the part of the code that was taken
2609 from the Samba project. Fortunately, the assertion is in a very simple
2610 function, so I have fixed this by reproducing the function inline in the
2611 one place where it is called, and arranging for authentication to fail
2612 instead of killing the process with assert().
2614 75. The SPA client code was not working when the server requested OEM rather
2615 than Unicode encoding.
2617 76. Added code to make require_files with a specific uid setting more usable in
2618 the case where statting the file as root fails - usually a non-root-mounted
2619 NFS file system. When this happens and the failure is EACCES, Exim now
2620 forks a subprocess and does the per-uid checking as the relevant uid.
2622 77. Added process_log_path.
2624 78. If log_file_path was not explicitly set, a setting of check_log_space or
2625 check_log_inodes was ignored.
2627 79. If a space check for the spool or log partitions fails, the incident is now
2628 logged. Of course, in the latter case the data may get lost...
2630 80. Added the %p formatting code to string_format() so that it can be used to
2631 print addresses in debug_print(). Adjusted all the address printing in the
2632 debugging in store.c to use %p rather than %d.
2634 81. There was a concern that a line of code in smtp_in.c could overflow a
2635 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2636 initially expressed, the concern was not well-founded, because trailing
2637 spaces are removed early. However, if the trailing spaces were followed by
2638 a NULL, they did not get removed, so the overflow was possible. Two fixes
2641 (a) I re-wrote the offending code in a cleaner fashion.
2642 (b) If an incoming SMTP command contains a NULL character, it is rejected
2645 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2646 runs initgroups(), so that if the Exim user is in any additional groups,
2647 they will be used during message reception.
2653 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob