1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.91 2005/03/15 14:09:12 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 TK/01. Added Yahoo DomainKeys support via libdomainkeys. See
11 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
13 TK/02. Fix ACL "control" statment not being available in MIME ACL.
15 TK/03. Fix ACL "regex" condition not being available in MIME ACL.
17 PH/01. Installed a patch from the Sieve maintainer that allows -bf to be used
18 to test Sieve filters that use "vacation".
20 PH/02. Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
21 that changes the way the GnuTLS parameters are stored in the cache file.
22 The new format can be generated externally. For backward compatibility,
23 if the data in the cache doesn't make sense, Exim assumes it has read an
24 old-format file, and it generates new data and writes a new file. This
25 means that you can't go back to an older release without removing the
28 PH/03. A redirect router that has both "unseen" and "one_time" set does not
29 work if there are any delivery delays because "one_time" forces the
30 parent to be marked "delivered", so its unseen clone is never tried
31 again. For this reason, Exim now forbids the simultaneous setting of
34 PH/04. Change 4.11/85 fixed an obscure bug concerned with addresses that are
35 redirected to themselves ("homonym" addresses). Read the long ChangeLog
36 entry if you want to know the details. The fix, however, neglected to
37 consider the case when local delivery batching is involved. The test for
38 "previously delivered" was not happening when checking to see if an
39 address could be batched with a previous (undelivered) one; under
40 certain circumstances this could lead to multiple deliveries to the same
41 address. A one-line patch to add the appropriate test fixes the bug.
43 PH/05. Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
44 in its include files, and this causes problems building Exim.
47 A note about Exim versions 4.44 and 4.50
48 ----------------------------------------
50 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
51 changes of various kinds. As a consequence, a big documentation update was
52 needed. This delayed the release for rather longer than seemed good, especially
53 in the light of a couple of (minor) security issues. Therefore, the changes
54 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
55 release. So 4.44 and 4.50 are in effect two different branches that both start
58 I have left the 4.50 change log unchanged; it contains all the changes since
59 4.43. The change log for 4.44 is below; many of its items are identical to
60 those for 4.50. This seems to be the most sensible way to preserve the
61 historical information.
67 1. Minor wording change to the doc/README.SIEVE file.
69 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
70 computation of the current number of files was incorrect.
72 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
73 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
76 4. Give more explanation in the error message when the command for a transport
77 filter fails to execute.
79 5. There are several places where Exim runs a non-Exim command in a
80 subprocess. The SIGUSR1 signal should be disabled for these processes. This
81 was being done only for the command run by the queryprogram router. It is
82 now done for all such subprocesses. The other cases are: ${run, transport
83 filters, and the commands run by the lmtp and pipe transports.
85 6. Added CONFIGURE_GROUP build-time option.
87 7. Some older OS have a limit of 256 on the maximum number of file
88 descriptors. Exim was using setrlimit() to set 1000 as a large value
89 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
90 systems. I've change it so that if it can't get 1000, it tries for 256.
92 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
93 was an oversight, and furthermore, ever since the addition of extra
94 controls (e.g. 4.43/32), the checks on when to allow different forms of
95 "control" were broken. There should now be diagnostics for all cases when a
96 control that does not make sense is encountered.
98 9. Added the /retain_sender option to "control=submission".
100 10. $recipients is now available in the predata ACL (oversight).
102 11. Tidy the search cache before the fork to do a delivery from a message
103 received from the command line. Otherwise the child will trigger a lookup
104 failure and thereby defer the delivery if it tries to use (for example) a
105 cached ldap connection that the parent has called unbind on.
107 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
108 of $address_data from the recipient verification was clobbered by the
111 13. The value of address_data from a sender verification is now available in
112 $sender_address_data in subsequent conditions in the ACL statement.
114 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
116 15. Added a new option "connect=<time>" to callout options, to set a different
119 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
120 was its contents. (It was OK if the option was not defined at all.)
122 17. A "Completed" log line is now written for messages that are removed from
123 the spool by the -Mrm option.
125 18. New variables $sender_verify_failure and $recipient_verify_failure contain
126 information about exactly what failed.
128 19. Added -dd to debug only the daemon process.
130 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
131 handles timeouts, both on the server side and network timeouts. Renamed the
132 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
134 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
135 It is now set to "smtps".
137 22. $host_address is now set to the target address during the checking of
140 23. When checking ignore_target_hosts for an ipliteral router, no host name was
141 being passed; this would have caused $sender_host_name to have been used if
142 matching the list had actually called for a host name (not very likely,
143 since this list is usually IP addresses). A host name is now passed as
146 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
147 code that specifies a non-restarting handler (typically sigaction() in
148 modern systems) in an attempt to fix a rare and obscure crash bug.
150 25. Narrowed the window for a race in the daemon that could cause it to ignore
151 SIGCHLD signals. This is not a major problem, because they are used only to
152 wake it up if nothing else does.
154 26. A malformed maildirsize file could cause Exim to calculate negative values
155 for the mailbox size or file count. Odd effects could occur as a result.
156 The maildirsize information is now recalculated if the size or filecount
159 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
160 support for a long time. Removed HAVE_SYS_VFS_H.
162 28. Installed the latest version of exipick from John Jetmore.
164 29. In an address list, if the pattern was not a regular expression, an empty
165 subject address (from a bounce message) matched only if the pattern was an
166 empty string. Non-empty patterns were not even tested. This was the wrong
167 because it is perfectly reasonable to use an empty address as part of a
168 database query. An empty address is now tested by patterns that are
169 lookups. However, all the other forms of pattern expect the subject to
170 contain a local part and a domain, and therefore, for them, an empty
171 address still always fails if the pattern is not itself empty.
173 30. Exim went into a mad DNS loop when attempting to do a callout where the
174 host was specified on an smtp transport, and looking it up yielded more
177 31. Re-factored the code for checking spool and log partition space into a
178 function that finds that data and another that does the check. The former
179 is then used to implement four new variables: $spool_space, $log_space,
180 $spool_inodes, and $log_inodes.
182 32. The RFC2047 encoding function was originally intended for short strings
183 such as real names; it was not keeping to the 75-character limit for
184 encoded words that the RFC imposes. It now respects the limit, and
185 generates multiple encoded words if necessary. To be on the safe side, I
186 have increased the buffer size for the ${rfc2047: expansion operator from
189 33. It is now permitted to omit both strings after an "if" condition; if the
190 condition is true, the result is "true". As before, when the second string
191 is omitted, a false condition yields an empty string. This makes it less
192 cumbersome to write custom ACL and router conditions.
194 34. Failure to deliver a bounce message always caused it to be frozen, even if
195 there was an errors_to setting on the router. The errors_to setting is now
198 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
199 canonical form (fully expanded) before being placed in
200 $sender_host_address.
202 36. The table in the code that translates DNS record types into text (T_A to
203 "A" for instance) was missing entries for NS and CNAME. It is just possible
204 that this could have caused confusion if both these types were looked up
205 for the same domain, because the text type is used as part of Exim's
206 per-process caching. But the chance of anyone hitting this buglet seems
209 37. The dnsdb lookup has been extended in a number of ways.
211 (1) There is a new type, "zns", which walks up the domain tree until it
212 finds some nameserver records. It should be used with care.
214 (2) There is a new type, "mxh", which is like "mx" except that it returns
215 just the host names, not the priorities.
217 (3) It is now possible to give a list of domains (or IP addresses) to be
218 looked up. The behaviour when one of the lookups defers can be
219 controlled by a keyword.
221 (4) It is now possible to specify the separator character for use when
222 multiple records are returned.
224 38. The dnslists ACL condition has been extended: it is now possible to supply
225 a list of IP addresses and/or domains to be looked up in a particular DNS
228 39. Added log_selector=+queue_time_overall.
230 40. When running the queue in the test harness, wait just a tad after forking a
231 delivery process, to get repeatability of debugging output.
233 41. Include certificate and key file names in error message when GnuTLS fails
234 to set them up, because the GnuTLS error message doesn't include the name
235 of the failing file when there is a problem reading it.
237 42. Allow both -bf and -bF in the same test run.
239 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
241 44. The "Exiscan patch" is now merged into the mainline Exim source.
243 45. Sometimes the final signoff response after QUIT could fail to get
244 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
245 < 0 before doing a fflush(). This bug looks as though it goes back to the
246 introduction of TLS in release 3.20, but "sometimes" must have been rare
247 because the tests only now provoked it.
249 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
250 (this can affect the format of dates).
252 47. exim_tidydb, when checking for the continued existence of a message for
253 which it has found a message-specific retry record, was not finding
254 messages that were in split spool directories. Consequently, it was
255 deleting retry records that should have stayed in existence.
257 48. Steve fixed some bugs in eximstats.
259 49. The SPA authentication driver was not abandoning authentication and moving
260 on to the next authenticator when an expansion was forced to fail,
261 contradicting the general specification for all authenticators. Instead it
262 was generating a temporary error. It now behaves as specified.
264 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
265 (the order specifies the preference for clients). The order is now AES256,
266 AES128, 3DES, ARCFOUR128.
268 51. Small patch to Sieve code - explicitly set From: when generating an
271 52. Exim crashed if a remote delivery caused a very long error message to be
272 recorded - for instance if somebody sent an entire SpamAssassin report back
273 as a large number of 550 error lines. This bug was coincidentally fixed by
274 increasing the size of one of Exim's internal buffers (big_buffer) that
275 happened as part of the Exiscan merge. However, to be on the safe side, I
276 have made the code more robust (and fixed the comments that describe what
279 53. Now that there can be additional text after "Completed" in log lines (if
280 the queue_time_overall log selector is set), a one-byte patch to exigrep
281 was needed to allow it to recognize "Completed" as not the last thing in
284 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
285 patch that reportedly fixes this has been added. I am not expert enough to
286 create a test for it. This is what the patch creator wrote:
288 "I found a little strange behaviour of ldap code when working with
289 Windows 2003 AD Domain, where users was placed in more than one
290 Organization Units. When I tried to give exim partial DN, the exit code
291 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
292 But simultaneously result of request was absolutely normal ldap result,
293 so I produce this patch..."
295 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
296 REFERENCE, so I have modified the code to exclude the patch when that macro
299 55. Some experimental protocols are using DNS PTR records for new purposes. The
300 keys for these records are domain names, not reversed IP addresses. The
301 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
302 leaves it alone. Component reversal etc. now happens only for IP addresses.
305 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
307 57. Double the size of the debug message buffer (to 2048) so that more of very
308 long debug lines gets shown.
310 58. The exicyclog utility now does better if the number of log files to keep
311 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
313 59. Two changes related to the smtp_active_hostname option:
315 (1) $smtp_active_hostname is now available as a variable.
316 (2) The default for smtp_banner uses $smtp_active_hostname instead
317 of $primary_hostname.
319 60. The host_aton() function is supposed to be passed a string that is known
320 to be a valid IP address. However, in the case of IPv6 addresses, it was
321 not checking this. This is a hostage to fortune. Exim now panics and dies
322 if the condition is not met. A case was found where this could be provoked
323 from a dnsdb PTR lookup with an IPv6 address that had more than 8
324 components; fortuitously, this particular loophole had already been fixed
325 by change 4.50/55 above.
327 If there are any other similar loopholes, the new check in host_aton()
328 itself should stop them being exploited. The report I received stated that
329 data on the command line could provoke the exploit when Exim was running as
330 exim, but did not say which command line option was involved. All I could
331 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
335 61. There was a buffer overflow vulnerability in the SPA authentication code
336 (which came originally from the Samba project). I have added a test to the
337 spa_base64_to_bits() function which I hope fixes it.
340 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
341 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
343 63. The daemon start-up calls getloadavg() while still root for those OS that
344 need the first call to be done as root, but it missed one case: when
345 deliver_queue_load_max is set with deliver_drop_privilege. This is
346 necessary for the benefit of the queue runner, because there is no re-exec
347 when deliver_drop_privilege is set.
349 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
352 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
353 was in use, was not putting the data itself into the right store pool;
354 consequently, it could be overwritten for a subsequent message in the same
355 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
358 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
360 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
361 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
362 the function were treating the return as a boolean value, which happened to
363 work because 0=false and not-0=true, but is not correct code.
365 68. The host_aton() function was not handling scoped IPv6 addresses (those
366 with, for example, "%eth0" on the end) correctly.
368 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
369 negated items (that is, ~something) in unsigned ints. Some compilers
370 apparently mutter when there is no cast.
372 70. If an address verification called from an ACL failed, and did not produce a
373 user-specific message (i.e. there was only a "system" message), nothing was
374 put in $acl_verify_message. In this situation, it now puts the system
377 71. Change 4.23/11 added synchronization checking at the start of an SMTP
378 session; change 4.31/43 added the unwanted input to the log line - except
379 that it did not do this in the start of session case. It now does.
381 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
382 This is wrong and can cause the other end to generate a synchronization
383 error if it is another Exim or anything else that does the synchronization
384 check. A QUIT command is no longer sent after a timeout.
386 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
389 74. The defer_ok option of callout verification was not working if it was used
390 when verifying addresses in header lines, that is, for this case:
392 verify = header_sender/callout=defer_ok
394 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
395 those file descriptors could be used for SMTP connections. If anything
396 wrote to stderr (the example that came up was "warn" in embedded Perl), it
397 could be sent to the SMTP client, causing chaos. The daemon now opens
398 stdin, stdout, and stderr to /dev/null when it puts itself into the
401 76. Arrange for output from Perl's "warn" command to be written to Exim's main
402 log by default. The user can override this with suitable Perl magic.
404 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
405 the log message when discard triggers, was not working for the DATA ACL or
406 for the non-SMTP ACL.
408 78. Error message wording change in sieve.c.
410 79. If smtp_accept_max_per_host was set, the number of connections could be
411 restricted to fewer than expected, because the daemon was trying to set up
412 a new connection before checking whether the processes handling previous
413 connections had finished. The check for completed processes is now done
414 earlier. On busy systems, this bug wouldn't be noticed because something
415 else would have woken the daemon, and it would have reaped the completed
418 80. If a message was submitted locally by a user whose login name contained one
419 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
420 It caused a spool format error. I have fixed the spool reading code. A
421 related problem was that the "from" clause in the Received: line became
422 illegal because of the space(s). It is now covered by ${quote_local_part.
424 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
427 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
429 83. Patch from Sieve maintainer to fix unterminated string problem in
432 84. Some minor changes to the Linux configuration files to help with other
433 OS variants using glibc.
435 85. One more patch for Sieve to update vacation handling to latest spec.
438 ----------------------------------------------------
439 See the note above about the 4.44 and 4.50 releases.
440 ----------------------------------------------------
446 1. Change 4.43/35 introduced a bug that caused file counts to be
447 incorrectly computed when quota_filecount was set in an appendfile
450 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
451 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
454 3. Give more explanation in the error message when the command for a transport
455 filter fails to execute.
457 4. There are several places where Exim runs a non-Exim command in a
458 subprocess. The SIGUSR1 signal should be disabled for these processes. This
459 was being done only for the command run by the queryprogram router. It is
460 now done for all such subprocesses. The other cases are: ${run, transport
461 filters, and the commands run by the lmtp and pipe transports.
463 5. Some older OS have a limit of 256 on the maximum number of file
464 descriptors. Exim was using setrlimit() to set 1000 as a large value
465 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
466 systems. I've change it so that if it can't get 1000, it tries for 256.
468 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
469 was an oversight, and furthermore, ever since the addition of extra
470 controls (e.g. 4.43/32), the checks on when to allow different forms of
471 "control" were broken. There should now be diagnostics for all cases when a
472 control that does not make sense is encountered.
474 7. $recipients is now available in the predata ACL (oversight).
476 8. Tidy the search cache before the fork to do a delivery from a message
477 received from the command line. Otherwise the child will trigger a lookup
478 failure and thereby defer the delivery if it tries to use (for example) a
479 cached ldap connection that the parent has called unbind on.
481 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
482 of $address_data from the recipient verification was clobbered by the
485 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
486 was its contents. (It was OK if the option was not defined at all.)
488 11. A "Completed" log line is now written for messages that are removed from
489 the spool by the -Mrm option.
491 12. $host_address is now set to the target address during the checking of
494 13. When checking ignore_target_hosts for an ipliteral router, no host name was
495 being passed; this would have caused $sender_host_name to have been used if
496 matching the list had actually called for a host name (not very likely,
497 since this list is usually IP addresses). A host name is now passed as
500 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
501 code that specifies a non-restarting handler (typically sigaction() in
502 modern systems) in an attempt to fix a rare and obscure crash bug.
504 15. Narrowed the window for a race in the daemon that could cause it to ignore
505 SIGCHLD signals. This is not a major problem, because they are used only to
506 wake it up if nothing else does.
508 16. A malformed maildirsize file could cause Exim to calculate negative values
509 for the mailbox size or file count. Odd effects could occur as a result.
510 The maildirsize information is now recalculated if the size or filecount
513 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
514 support for a long time. Removed HAVE_SYS_VFS_H.
516 18. Updated exipick to current release from John Jetmore.
518 19. Allow an empty sender to be matched against a lookup in an address list.
519 Previously the only cases considered were a regular expression, or an
522 20. Exim went into a mad DNS lookup loop when doing a callout where the
523 host was specified on the transport, if the DNS lookup yielded more than
526 21. The RFC2047 encoding function was originally intended for short strings
527 such as real names; it was not keeping to the 75-character limit for
528 encoded words that the RFC imposes. It now respects the limit, and
529 generates multiple encoded words if necessary. To be on the safe side, I
530 have increased the buffer size for the ${rfc2047: expansion operator from
533 22. Failure to deliver a bounce message always caused it to be frozen, even if
534 there was an errors_to setting on the router. The errors_to setting is now
537 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
538 canonical form (fully expanded) before being placed in
539 $sender_host_address.
541 24. Updated eximstats to version 1.33
543 25. Include certificate and key file names in error message when GnuTLS fails
544 to set them up, because the GnuTLS error message doesn't include the name
545 of the failing file when there is a problem reading it.
547 26. Expand error message when OpenSSL has problems setting up cert/key files.
550 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
551 (this can affect the format of dates).
553 28. exim_tidydb, when checking for the continued existence of a message for
554 which it has found a message-specific retry record, was not finding
555 messages that were in split spool directories. Consequently, it was
556 deleting retry records that should have stayed in existence.
558 29. eximstats updated to version 1.35
559 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
560 1.35 - bugfix such that pie charts by volume are generated correctly
562 30. The SPA authentication driver was not abandoning authentication and moving
563 on to the next authenticator when an expansion was forced to fail,
564 contradicting the general specification for all authenticators. Instead it
565 was generating a temporary error. It now behaves as specified.
567 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
568 (the order specifies the preference for clients). The order is now AES256,
569 AES128, 3DES, ARCFOUR128.
571 31. Small patch to Sieve code - explicitly set From: when generating an
574 32. Exim crashed if a remote delivery caused a very long error message to be
575 recorded - for instance if somebody sent an entire SpamAssassin report back
576 as a large number of 550 error lines. This bug was coincidentally fixed by
577 increasing the size of one of Exim's internal buffers (big_buffer) that
578 happened as part of the Exiscan merge. However, to be on the safe side, I
579 have made the code more robust (and fixed the comments that describe what
582 33. Some experimental protocols are using DNS PTR records for new purposes. The
583 keys for these records are domain names, not reversed IP addresses. The
584 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
585 leaves it alone. Component reversal etc. now happens only for IP addresses.
588 34. The host_aton() function is supposed to be passed a string that is known
589 to be a valid IP address. However, in the case of IPv6 addresses, it was
590 not checking this. This is a hostage to fortune. Exim now panics and dies
591 if the condition is not met. A case was found where this could be provoked
592 from a dnsdb PTR lookup with an IPv6 address that had more than 8
593 components; fortuitously, this particular loophole had already been fixed
594 by change 4.50/55 or 4.44/33 above.
596 If there are any other similar loopholes, the new check in host_aton()
597 itself should stop them being exploited. The report I received stated that
598 data on the command line could provoke the exploit when Exim was running as
599 exim, but did not say which command line option was involved. All I could
600 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
604 35. There was a buffer overflow vulnerability in the SPA authentication code
605 (which came originally from the Samba project). I have added a test to the
606 spa_base64_to_bits() function which I hope fixes it.
609 36. The daemon start-up calls getloadavg() while still root for those OS that
610 need the first call to be done as root, but it missed one case: when
611 deliver_queue_load_max is set with deliver_drop_privilege. This is
612 necessary for the benefit of the queue runner, because there is no re-exec
613 when deliver_drop_privilege is set.
615 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
616 was in use, was not putting the data itself into the right store pool;
617 consequently, it could be overwritten for a subsequent message in the same
618 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
621 38. Sometimes the final signoff response after QUIT could fail to get
622 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
623 < 0 before doing a fflush(). This bug looks as though it goes back to the
624 introduction of TLS in release 3.20, but "sometimes" must have been rare
625 because the tests only now provoked it.
631 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
632 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
633 Now it returns an integer. A number of calls were still expecting a T/F
634 return. Fortuitously, in all cases, the tests worked in OK situations,
635 which is the norm. However, things would have gone wrong on any write
636 failures on the smtp file descriptor. This function is used when sending
637 messages over SMTP and also when doing verify callouts.
639 2. When Exim is called to do synchronous delivery of a locally submitted
640 message (the -odf or -odi options), it no longer closes stderr before doing
643 3. Implemented the mua_wrapper option.
645 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
647 5. Implemented the functions header_remove(), header_testname(),
648 header_add_at_position(), and receive_remove_recipient(), and exported them
651 6. If an ACL "warn" statement specified the addition of headers, Exim already
652 inserted X-ACL-Warn: at the start if there was no header name. However, it
653 was not making this test for the second and subsequent header lines if
654 there were newlines in the string. This meant that an invalid header could
655 be inserted if Exim was badly configured.
657 7. Allow an ACL "warn" statement to add header lines at the start or after all
658 the Received: headers, as well as at the end.
660 8. Added the rcpt_4xx retry error code.
662 9. Added postmaster_mailfrom=xxx to callout verification option.
664 10. Added mailfrom=xxxx to the callout verification option, for verify=
667 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
668 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
670 12. Inserted some casts to stop certain compilers warning when using pointer
671 differences as field lengths or precisions in printf-type calls (mostly
672 affecting debugging statements).
674 13. Added optional readline() support for -be (dynamically loaded).
676 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
677 same clock tick as a message's arrival, so that its received time was the
678 same as the "first fail" time on the retry record, and that message
679 remained on the queue past the ultimate address timeout, every queue runner
680 would try a delivery (because it was past the ultimate address timeout) but
681 after another failure, the ultimate address timeout, which should have then
682 bounced the address, did not kick in. This was a "< instead of <=" error;
683 in most cases the first failure would have been in the next clock tick
684 after the received time, and all would be well.
686 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
687 being recognized when the domain list was tested by the match_domain
688 condition in an expansion string.
690 16. Added the ${str2b64: operator.
692 17. Exim was always calling setrlimit() to set a large limit for the number of
693 processes, without checking whether the existing limit was already
694 adequate. (It did check for the limit on file descriptors.) Furthermore,
695 errors from getrlimit() and setrlimit() were being ignored. Now they are
696 logged to the main and panic logs, but Exim does carry on, to try to do its
697 job under whatever limits there are.
699 18. Imported PCRE 5.0.
701 19. Trivial typo in log message " temporarily refused connection" (the leading
704 20. If the log selector return_path_on_delivery was set and an address was
705 redirected to /dev/null, the delivery process crashed because it assumed
706 that a return path would always be set for a "successful" delivery. In this
707 case, the whole delivery is bypassed as an optimization, and therefore no
710 21. Internal re-arrangement: the function for sending a challenge and reading
711 a response while authentication was assuming a zero-terminated challenge
712 string. It's now changed to take a pointer and a length, to allow for
713 binary data in such strings.
715 22. Added the cyrus_sasl authenticator (code supplied by MBM).
717 23. Exim was not respecting finduser_retries when seeking the login of the
718 uid under which it was called; it was always trying 10 times. (The default
719 setting of finduser_retries is zero.) Also, it was sleeping after the final
720 failure, which is pointless.
722 24. Implemented tls_on_connect_ports.
724 25. Implemented acl_smtp_predata.
726 26. If the domain in control=submission is set empty, Exim assumes that the
727 authenticated id is a complete email address when it generates From: or
728 Sender: header lines.
730 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
731 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
732 chown and chgrp in /bin and hostname in /usr/bin.
734 28. Exim was keeping the "process log" file open after each use, just as it
735 does for the main log. This opens the possibility of it remaining open for
736 long periods when the USR1 signal hits a daemon. Occasional processlog
737 errors were reported, that could have been caused by this. Anyway, it seems
738 much more sensible not to leave this file open at all, so that is what now
741 29. The long-running daemon process does not normally write to the log once it
742 has entered its main loop, and it closes the log before doing so. This is
743 so that log files can straightforwardly be renamed and moved. However,
744 there are a couple of unusual error situations where the daemon does write
745 log entries, and I had neglected to close the log afterwards.
747 30. The text of an SMTP error response that was received during a remote
748 delivery was being truncated at 512 bytes. This is too short for some of
749 the long messages that one sometimes sees. I've increased the limit to
752 31. It is now possible to make retry rules that apply only when a message has a
753 specific sender, in particular, an empty sender.
755 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
756 it possible to be selective about when SMTP synchronization is enforced.
758 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
760 32. Implemented hosts_connection_nolog.
762 33. Added an ACL for QUIT.
764 34. Setting "delay_warning=" to disable warnings was not working; it gave a
767 35. Added mailbox_size and mailbox_filecount to appendfile.
769 36. Added control = no_multiline_responses to ACLs.
771 37. There was a bug in the logic of the code that waits for the clock to tick
772 in the case where the clock went backwards by a substantial amount such
773 that the microsecond fraction of "now" was more than the microsecond
774 fraction of "then" (but the whole seconds number was less).
776 38. Added support for the libradius Radius client library this is found on
777 FreeBSD (previously only the radiusclient library was supported).
783 1. When certain lookups returned multiple values in the form name=value, the
784 quoting of the values was not always being done properly. Specifically:
785 (a) If the value started with a double quote, but contained no whitespace,
787 (b) If the value contained whitespace other than a space character (i.e.
788 tabs or newlines or carriage returns) it was not quoted.
789 This fix has been applied to the mysql and pgsql lookups by writing a
790 separate quoting function and calling it from the lookup code. The fix
791 should probably also be applied to nisplus, ibase and oracle lookups, but
792 since I cannot test any of those, I have not disturbed their existing code.
794 2. A hit in the callout cache for a specific address caused a log line with no
795 reason for rejecting RCPT. Now it says "Previous (cached) callout
796 verification failure".
798 3. There was an off-by-one bug in the queryprogram router. An over-long
799 return line was truncated at 256 instead of 255 characters, thereby
800 overflowing its buffer with the terminating zero. As well as fixing this, I
801 have increased the buffer size to 1024 (and made a note to document this).
803 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
804 when Exim is waiting for an SMTP response from a remote server, Exim
805 restarts its select() call on the socket, thereby resetting its timeout.
806 This is not a problem when such interrupts are rare. Somebody set up a cron
807 job to run exiwhat every 2 minutes, which is less than the normal select()
808 timeout (5 or 10 minutes). This meant that the select() timeout never
809 kicked in because it was always reset. I have fixed this by comparing the
810 time when an interrupt arrives with the time at the start of the first call
811 to select(). If more time than the timeout has elapsed, the interrupt is
812 treated as a timeout.
814 5. Some internal re-factoring in preparation for the addition of Sieve
815 extensions (by MH). In particular, the "personal" test is moved to a
816 separate function, and given an option for scanning Cc: and Bcc: (which is
817 not set for Exim filters).
819 6. When Exim created an email address using the login of the caller as the
820 local part (e.g. when creating a From: or Sender: header line), it was not
821 quoting the local part when it contained special characters such as @.
823 7. Installed new OpenBSD configuration files.
825 8. Reworded some messages for syntax errors in "and" and "or" conditions to
826 try to make them clearer.
828 9. Callout options, other than the timeout value, were being ignored when
829 verifying sender addresses in header lines. For example, when using
831 verify = header_sender/callout=no_cache
833 the cache was (incorrectly) being used.
835 10. Added a missing instance of ${EXE} to the exim_install script; this affects
836 only the Cygwin environment.
838 11. When return_path_on_delivery was set as a log selector, if different remote
839 addresses in the same message used different return paths and parallel
840 remote delivery occurred, the wrong values would sometimes be logged.
841 (Whenever a remote delivery process finished, the return path value from
842 the most recently started remote delivery process was logged.)
844 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
845 lines when AUTH and/or TLS are in use. This is the "received protocol"
846 field. Exim used to use "asmtp" for authenticated SMTP, without any
847 indication (in the protocol name) for TLS use. Now it follows the RFC and
848 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
849 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
850 names appear in log lines as well as in Received: header lines.
852 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
853 extensions, and comparison tests, and to fix some bugs.
855 14. Changes to the "personal" filter test:
857 (1) The test was buggy in that it was just doing the equivalent of
858 "contains" tests on header lines. For example, if a user's address was
859 anne@some.where, the "personal" test would incorrectly be true for
861 To: susanne@some.where
863 This test is now done by extracting each address from the header in turn,
864 and checking the entire address. Other tests that are part of "personal"
865 are now done using regular expressions (for example, to check local parts
866 of addresses in From: header lines).
868 (2) The list of non-personal local parts in From: addresses has been
869 extended to include "listserv", "majordomo", "*-request", and "owner-*",
870 taken from the Sieve specification recommendations.
872 (3) If the message contains any header line starting with "List-" it is
873 treated as non-personal.
875 (4) The test for "circular" in the Subject: header line has been removed
876 because it now seems ill-conceived.
878 15. Minor typos in src/EDITME comments corrected.
880 16. Installed latest exipick from John Jetmore.
882 17. If headers_add on a router specified a text string that was too long for
883 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
884 of string_sprintf() is now avoided.
886 18. $message_body_size was not set (it was always zero) when running the DATA
887 ACL and the local_scan() function.
889 19. For the "mail" command in an Exim filter, no default was being set for
890 the once_repeat time, causing a random time value to be used if "once" was
891 specified. (If the value happened to be <= 0, no repeat happened.) The
892 default is now 0s, meaning "never repeat". The "vacation" command was OK
893 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
894 (I found it when inspecting the code).
896 20. There is now an overall timeout for performing a callout verification. It
897 defaults to 4 times the callout timeout, which applies to individual SMTP
898 commands during the callout. The overall timeout applies when there is more
899 than one host that can be tried. The timeout is checked before trying the
900 next host. This prevents very long delays if there are a large number of
901 hosts and all are timing out (e.g. when the network connections are timing
902 out). The value of the overall timeout can be changed by specifying an
903 additional sub-option for "callout", called "maxwait". For example:
905 verify = sender/callout=5s,maxwait=20s
907 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
908 to the end before writing, but this should make it even safer).
910 22. Exim was forgetting that it had advertised PIPELINING for the second and
911 subsequent messages on an SMTP connection. It was also not resetting its
912 memory on STARTTLS and an internal HELO.
914 23. When Exim logs an SMTP synchronization error within a session, it now
915 records whether PIPELINING has been advertised or not.
917 24. Added 3 instances of "(long int)" casts to time_t variables that were being
918 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
919 rather than long int.
921 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
923 26. Added the never_mail option to autoreply.
929 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
930 crash if the getsockname() call failed; this can happen if a connection is
931 closed very soon after it is established. The problem was simply in the
932 order in which certain operations were done, causing Exim to try to write
933 to the SMTP stream before it had set up the file descriptor. The bug has
934 been fixed by making things happen in the correct order.
940 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
941 before the connection was closed, thus losing the rejection response.
943 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
944 some early Solaris releases, but causes trouble in current releases where
945 socklen_t is defined.
947 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
950 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
953 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
954 that releasing the top of it at the end releases what was used for sub-
955 expansions (unless the block got too big). However, discard this block if
956 the first thing is a variable or header, so that we can use its block when
957 it is dynamic (useful for very large $message_headers, for example).
959 6. Lookups now cache *every* query, not just the most recent. A new, separate
960 store pool is used for this. It can be recovered when all lookup caches are
961 flushed. Lookups now release memory at the end of their result strings.
962 This has involved some general refactoring of the lookup sources.
964 7. Some code has been added to the store_xxx() functions to reduce the amount
965 of flapping under certain conditions.
967 8. log_incoming_interface used to affect only the <= reception log lines. Now
968 it causes the local interface and port to be added to several more SMTP log
969 lines, for example "SMTP connection from", and rejection lines.
971 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
973 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
975 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
976 could be overwritten at the end of the current message (or the start of a
977 new message if it was set in a HELO ACL). The value is now preserved for
978 the duration of the SMTP connection.
980 12. If a transport had a headers_rewrite setting, and a matching header line
981 contained an unqualified address, that address was qualified, even if it
982 did not match any rewriting rules. The underlying bug was that the values
983 of the flags that permit the existence of unqualified sender and recipient
984 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
985 non-local messages, and by -bnq for local messages) were not being
986 preserved with the message after it was received.
988 13. When Exim was logging an SMTP synchronization error, it could sometimes log
989 "next input=" as part of the text comprising the host identity instead of
990 the correct text. The code was using the same buffer for two different
991 strings. However, depending on which order the printing function evaluated
992 its arguments, the bug did not always show up. Under Linux, for example, my
993 test suite worked just fine.
995 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
996 doesn't work with some older versions of Perl. It has been changed to "my",
997 which in any case is probably the better facility to use.
999 15. A really picky compiler found some instances of statements for creating
1000 error messages that either had too many or two few arguments for the format
1003 16. The size of the buffer for calls to the DNS resolver has been increased
1004 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1005 for addresses that have a lot of PTR records. This alleviates a problem; it
1006 does not fully solve it.
1008 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1009 buffer now truncates the list and logs the incident, which is the same
1010 action as happens when Exim is looking up a host name and its aliases.
1011 Previously in this situation something unpredictable would happen;
1012 sometimes it was "internal error: store_reset failed".
1014 18. If a server dropped the connection unexpectedly when an Exim client was
1015 using GnuTLS and trying to read a response, the client delivery process
1016 crashed while trying to generate an error log message.
1018 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1019 string, for example:
1021 warn message = H1: something\nH2: something
1023 the text was added as a single header line from Exim's point of view
1024 though it ended up OK in the delivered message. However, searching for the
1025 second and subsequent header lines using $h_h2: did not work. This has been
1026 fixed. Similarly, if a system filter added multiple headers in this way,
1027 the routers could not see them.
1029 20. Expanded the error message when iplsearch is called with an invalid key to
1030 suggest using net-iplsearch in a host list.
1032 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1033 ACLs are no longer actually imposed (and a message to that effect is
1036 22. If a "gecos" field in a passwd entry contained escaped characters, in
1037 particular, if it contained a \" sequence, Exim got it wrong when building
1038 a From: or a Sender: header from that name. A second bug also caused
1039 incorrect handling when an unquoted " was present following a character
1040 that needed quoting.
1042 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1043 was not being matched caselessly.
1045 24. Arranged for all hyphens in the exim.8 source to be escaped with
1048 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1049 properly. Recipient callout cache records were still being keyed to include
1050 the sender, even when use_sender was set false. This led to far more
1051 callouts that were necessary. The sender is no longer included in the key
1052 when use_sender is false.
1054 26. Added "control = submission" modifier to ACLs.
1056 27. Added the ${base62d: operator to decode base 62 numbers.
1058 28. dnsdb lookups can now access SRV records.
1060 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1061 the configuration file.
1063 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1064 (-v) mode. This makes the output for a verbose queue run more intelligible.
1066 31. Added a use_postmaster feature to recipient callouts.
1068 32. Added the $body_zerocount variable, containing the number of binary zero
1069 bytes in the message body.
1071 33. The time of last modification of the "new" subdirectory is now used as the
1072 "mailbox time last read" when there is a quota error for a maildir
1075 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1077 35. Added +ignore_unknown as a special item in host lists.
1079 36. Code for decoding IPv6 addresses in host lists is now included, even if
1080 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1081 address was recognized as an IP address, but was then not correctly decoded
1082 into binary, causing unexpected and incorrect effects when compared with
1089 1. Very minor rewording of debugging text in manualroute to say "list of
1090 hosts" instead of "hostlist".
1092 2. If verify=header_syntax was set, and a header line with an unqualified
1093 address (no domain) and a large number of spaces between the end of the
1094 name and the colon was received, the reception process suffered a buffer
1095 overflow, and (when I tested it) crashed. This was caused by some obsolete
1096 code that should have been removed. The fix is to remove it!
1098 3. When running in the test harness, delay a bit after writing a bounce
1099 message to get a bit more predictability in the log output.
1101 4. Added a call to search_tidyup() just before forking a reception process. In
1102 theory, someone could use a lookup in the expansion of smtp_accept_max_
1103 per_host which, without the tidyup, could leave open a database connection.
1105 5. Added the variables $recipient_data and $sender_data which get set from a
1106 lookup success in an ACL "recipients" or "senders" condition, or a router
1107 "senders" option, similar to $domain_data and $local_part_data.
1109 6. Moved the writing of debug_print from before to after the "senders" test
1112 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1113 problems for message scanning, either using a data ACL, or using
1114 local_scan() because the Received: header was not generated till after they
1115 were called (in order to set the time as the time of reception completion).
1116 I have revised the way this works. The header is now generated after the
1117 body is received, but before the ACL or local_scan() are called. After they
1118 are run, the timestamp in the header is updated.
1124 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1125 before starting a queue runner without re-exec. This happened only when
1126 deliver_drop_privilege was set or when the Exim user was set to root. The
1127 effect of the bug was that timeouts during subsequent deliveries caused
1128 crashes instead of being properly handled. The handler is now left at its
1129 default (and expected) setting.
1131 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1132 message, again when deliver_drop_privilege is set or Exim is run as root.
1133 The bug described in (1) was not present in this case, but the tidying up
1134 of the other signals was missing. I have made the two cases consistent.
1136 3. The ignore_target_hosts setting on a manualroute router was being ignored
1137 for hosts that were looked up using the /MX notation.
1139 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1142 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1143 operated on the sender address. After changing the $sender_address to <>
1144 for the sender address verify, Exim was re-instated it as the original
1145 (before rewriting) address, but remembering that it had rewritten it, so it
1146 wasn't rewriting it again. This bug also had the effect of breaking the
1147 sender address verification caching when the sender address was rewritten.
1149 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1150 This has been changed so that if the ip literal address matches
1151 ignore_target_hosts, the router declines.
1153 7. Added expansion conditions match_domain, match_address, and match_local_
1154 part (NOT match_host).
1156 8. The placeholder for the Received: header didn't have a length field set.
1158 9. Added code to Exim itself and to exim_lock to test for a specific race
1159 condition that could lead to file corruption when using MBX delivery. The
1160 issue is with the lockfile that is created in /tmp. If this file is removed
1161 after a process has opened it but before that process has acquired a lock,
1162 there is the potential for a second process to recreate the file and also
1163 acquire a lock. This could lead to two Exim processes writing to the file
1164 at the same time. The added code performs the same test as UW imapd; it
1165 checks after acquiring the lock that its file descriptor still refers to
1166 the same named file.
1168 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1169 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1172 11. Added the smtp_active_hostname option. If used, this will typically be made
1173 to depend on the incoming interface address. Because $interface_address is
1174 not set up until the daemon has forked a reception process, error responses
1175 that can happen earlier (such as "too many connections") no longer contain
1178 12. If an expansion in a condition on a "warn" statement fails because a lookup
1179 defers, the "warn" statement is abandoned, and the next ACL statement is
1180 processed. Previously this caused the whole ACL to be aborted.
1182 13. Added the iplsearch lookup type.
1184 14. Added ident_timeout as a log selector.
1186 15. Added tls_certificate_verified as a log selector.
1188 16. Added a global option tls_require_ciphers (compare the smtp transport
1189 option of the same name). This controls incoming TLS connections.
1191 17. I finally figured out how to make tls_require_ciphers do a similar thing
1192 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1193 before starting the TLS session.
1195 18. Tabs are now shown as \t in -bP output.
1197 19. If the log selector return_path_on_delivery was set, Exim crashed when
1198 bouncing a message because it had too many Received: header lines.
1200 20. If two routers both had headers_remove settings, and the first one included
1201 a superfluous trailing colon, the final name in the first list and the
1202 first name in the second list were incorrectly joined into one item (with a
1203 colon in the middle).
1209 1. Added -C and -D options to the exinext utility, mainly to make it easier
1210 to include in the automated testing, but these could be helpful when
1211 multiple configurations are in use.
1213 2. The exinext utility was not formatting the output nicely when there was
1214 an alternate port involved in the retry record key, nor when there was a
1215 message id as well (for retries that were specific to a specific message
1216 and a specific host). It was also confused by IPv6 addresses, because of
1217 the additional colons they contain. I have fixed the IPv4 problem, and
1218 patched it up to do a reasonable job for IPv6.
1220 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1221 delivery, the log line now contains "pipelined" if PIPELINING was used.
1223 4. An SMTP transport process used to panic and die if the bind() call to set
1224 an explicit outgoing interface failed. This has been changed; it is now
1225 treated in the same way as a connect() failure.
1227 5. A reference to $sender_host_name in the part of a conditional expansion
1228 that was being skipped was still causing a DNS lookup. This no longer
1231 6. The def: expansion condition was not recognizing references to header lines
1232 that used bh_ and bheader_.
1234 7. Added the _cache feature to named lists.
1236 8. The code for checking quota_filecount in the appendfile transport was
1237 allowing one more file than it should have been.
1239 9. For compatibility with Sendmail, the command line option
1247 and sets the incoming protocol and host name (for trusted callers). The
1248 host name and its colon can be omitted when only the protocol is to be set.
1249 Note the Exim already has two private options, -pd and -ps, that refer to
1250 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1251 "s", but I don't think that's a major issue.
1253 10. A number of refactoring changes to the code, none of which should affect
1256 (a) The number of logging options was getting close to filling up the
1257 32-bit word that was used as a bit map. I have split them into two classes:
1258 those that are passed in the argument to log_write(), and those that are
1259 only ever tested independently outside of that function. These are now in
1260 separate 32-bit words, so there is plenty of room for expansion again.
1261 There is no change in the user interface or the logging behaviour.
1263 (b) When building, for example, log lines, the code previously used a
1264 macro that called string_cat() twice, in order to add two strings. This is
1265 not really sufficiently general. Furthermore, there was one instance where
1266 it was actually wrong because one of the argument was used twice, and in
1267 one call a function was used. (As it happened, calling the function twice
1268 did not affect the overall behaviour.) The macro has been replaced by a
1269 function that can join an arbitrary number of extra strings onto a growing
1272 (c) The code for expansion conditions now uses a table and a binary chop
1273 instead of a serial search (which was left over from when there were very
1274 few conditions). Also, it now recognizes conditions like "pam" even when
1275 the relevant support is not compiled in: a suitably worded error message is
1276 given if an attempt is made to use such a condition.
1278 11. Added ${time_interval:xxxxx}.
1280 12. A bug was causing one of the ddress fields not to be passed back correctly
1281 from remote delivery subprocesses. The field in question was not being
1282 subsequently used, so this caused to problems in practice.
1284 13. Added new log selectors queue_time and deliver_time.
1286 14. Might have fixed a bug in maildirsizefile handling that threw up
1287 "unexpected character" debug warnings, and recalculated the data
1288 unnecessarily. In any case, I expanded the warning message to give more
1291 15. Added the message "Restricted characters in address" to the statements in
1292 the default ACL that block characters like @ and % in local parts.
1294 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1295 Three changes have been made:
1297 (a) There was a serious bug; a negative response to MAIL caused the whole
1298 recipient domain to be cached as invalid, thereby blocking all messages
1299 to all local parts at the same domain, from all senders. This bug has
1300 been fixed. The domain is no longer cached after a negative response to
1301 MAIL if the sender used is not empty.
1303 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1306 (c) A new callout option, "use_sender" has been added for people who want
1307 the modified behaviour.
1313 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1316 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1317 indeed breaks things for older releases.
1319 3. Added additional logging to the case where there is a problem reading data
1320 from a filter that is running in a subprocess using a pipe, in order to
1321 try to track down a specific problem.
1323 4. Testing facility fudge: when running in the test harness and attempting
1324 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1325 getting "No route to host". Convert this to a timeout.
1327 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1330 6. Some OS don't have socklen_t but use size_t instead. This affects the
1331 fifth argument of getsockopt() amongst other things. This is now
1332 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1333 can be set for individual OS. I have set it for SunOS5, OSF1, and
1334 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1335 some earlier ones do not.
1337 7. Change 4.30/15 was not doing the test caselessly.
1339 8. The standard form for an IPv6 address literal was being rejected by address
1340 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1341 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1342 this, as well as the form without the "IPv6" on the front (but only when
1343 address literals are enabled, of course).
1345 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1347 10. Exim crashed if a message with an empty sender address specified by -f
1348 encountered a router with an errors_to setting. This could be provoked only
1349 by a command such as
1353 where an empty string was supplied; "<>" did not hit this bug.
1355 11. Installed PCRE release 4.5.
1357 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1358 remained set. It is now erased.
1360 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1361 times from message ids (which are base 36 rather than the normal 62).
1363 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1364 were being counted as actual protocol errors, and logged if the log
1365 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1366 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1367 and DATA following a set of rejected RCPTs do not count as protocol errors.
1368 In other words, Exim assumes they were pipelined, though this may not
1369 actually be the case. Of course, in all cases the client gets an
1370 appropriate error code.
1372 15. If a lookup fails in an ACL condition, a message about the failure may
1373 be available; it is used if testing the ACL cannot continue, because most
1374 such messages specify what the cause of the deferral is. However, some
1375 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1376 that caused an old message to be retained and used if a later statement
1377 caused a defer, replacing the real cause of the deferral.
1379 16. If an IP address had so many PTR records that the DNS lookup buffer
1380 was not large enough to hold them, Exim could crash while trying to process
1381 the truncated data. It now detects and logs this case.
1383 17. Further to 4.21/58, another change has been made: if (and only if) the
1384 first line of a message (the first header line) ends with CRLF, a bare LF
1385 in a subsequent header line has a space inserted after it, so as not to
1386 terminate the header.
1388 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1389 unnecessarily, used atoi() instead of strtol(), and didn't check the
1390 termination when getting file sizes from file names by regex.
1392 19. Completely re-implemented the support for maildirsize files, in the light
1393 of a number of problems with the previous contributed implementation
1394 (4.30/29). In particular:
1396 . If the quota is zero, the maildirsize file is maintained, but no quota is
1399 . If the maildir directory does not exist, it is created before any attempt
1400 to write a maildirsize file.
1402 . The quota value in the file is just a cache; if the quota is changed in
1403 the transport, the new value overrides.
1405 . A regular expression is available for excluding directories from the
1408 20. The autoreply transport checks the characters in options that define the
1409 message's headers; it allows continued headers, but it was checking with
1410 isspace() after an embedded newline instead of explicitly looking for a
1413 21. If all the "regular" hosts to which an address was routed had passed their
1414 expiry times, and had not reached their retry times, the address was
1415 bounced, even if fallback hosts were defined. Now Exim should go on to try
1418 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1419 equivalent code in the SMTP transport. Some hosts send humungous responses
1420 to HELO/EHLO, more than 1024 it seems.
1422 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1423 gives compiler warnings in some environments. I've now done it "properly",
1426 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1427 (because of problems with the built-in one) was declared to return uschar *
1428 instead of char *, causing compiler failure.
1430 25. Fixed a file descriptor leak when processing alias/forward files.
1432 26. Fixed a minor format string issue in dbfn.c.
1434 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1436 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1437 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1440 29. When a sender address is verified, it is cached, to save repeating the test
1441 when there is more than one recipient in a message. However, when the
1442 verification involves a callout, it is possible for different callout
1443 options to be set for different recipients. It is too complicated to keep
1444 track of this in the cache, so now Exim always runs a verification when a
1445 callout is required, relying on the callout cache for the optimization.
1446 The overhead is duplication of the address routing, but this should not be
1449 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1450 to be verified with callout=postmaster, and the main callout worked but the
1451 postmaster check failed, the verification correctly failed. However, if a
1452 subsequent RCPT command asked for sender verification *without* the
1453 postmaster check, incorrect caching caused this verification also to fail,
1456 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1457 it was not caching the DNS options (qualify_single, search_parents) that
1458 were used when the lookup failed. A subsequent lookup with different
1459 options therefore always gave the same answer, though there were cases
1460 where it should not have. (Example: a "domains = !$mx_any" option on a
1461 dnslookup router: the "domains" option is always processed without any
1462 widening, but the router might have qualify_single set.) Now Exim uses the
1463 cached value only when the same options are set.
1465 32. Added John Jetmore's "exipick" utility to the distribution.
1467 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1468 than a timeout (e.g. a certificate is required, and is not provided), an
1469 Exim server now closes the connection immediately. Previously it waited for
1470 the client to close - but if the client is SSL, it seems that they each
1471 wait for each other, leading to a delay before one of them times out.
1473 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1474 maintained 0.8.x compatibility because I don't think many are using it, and
1475 it is clearly obsolete.
1477 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1480 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1481 client certificate was expired. A simple patch fixes this, though I don't
1482 understand the full logic of why the verify callback is called multiple
1485 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1486 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1487 which causes problems with some clients (such as the Certicom SSL Plus
1488 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1489 disables the coutermeasure allowing Eudora to connect."
1491 38. Exim was not checking that a write() to a log file succeeded. This could
1492 lead to Bad Things if a log got too big, in particular if it hit a file
1493 size limit. Exim now panics and dies if it cannot write to a log file, just
1494 as it does if it cannot open a log file.
1496 39. Modified OS/Makefile-Linux so that it now contains
1498 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1500 The two -D definitions ensure that Exim is compiled with large file
1501 support, which makes it possible to handle log files that are bigger than
1504 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1505 instance) a domain was checked against a named list that involved a lookup,
1506 causing $domain_data to be set, then another domain was checked against the
1507 same list, then the first domain was re-checked, the value of $domain_data
1508 after the final check could be wrong. In particular, if the second check
1509 failed, it could be set empty. This bug probably also applied to
1512 41. The strip_trailing_dot option was not being applied to the address given
1513 with the -f command-line option.
1515 42. The code for reading a message's header from the spool was incrementing
1516 $received_count, but never initializing it. This meant that the value was
1517 incorrect (doubled) while delivering a message in the same process in which
1518 it was received. In the most common configuration of Exim, this never
1519 happens - a fresh exec is done - but it can happen when
1520 deliver_drop_privilege is set.
1522 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1523 it now includes up to 150 characters of the unexpected data in the log
1526 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1527 and building data strings. The size of both of these buffers was 10 000
1528 bytes - far larger than anybody would *ever* want, thought I. Needless to
1529 say, somebody hit the limit. I have increased the maximum line length to
1530 20 000 and the maximum data length of concatenated lines to 100 000. I have
1531 also fixed two bugs, because there was no checking on these buffers. Tsk,
1532 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1533 buffer is too small.
1535 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1536 lsearch lookups. Now it does.
1538 46. When parsing a route_list item in a manualroute router, a fixed-length
1539 buffer was used for the list of hosts. I made this 1024 bytes long,
1540 thinking that nobody would ever have a list of hosts that long. Wrong.
1541 Somebody had a whole pile of complicated expansion conditions, and the
1542 string was silently truncated, leading to an expansion error. It turns out
1543 that it is easier to change to an unlimited length (owing to other changes
1544 that have happened since this code was originally written) than to build
1545 structure for giving a limitation error. The length of the item that
1546 expands into the list of hosts is now unlimited.
1548 47. The lsearch lookup could not handle data where the length of text line was
1549 more than 4095 characters. Such lines were truncated, leading to shortened
1550 data being returned. It should now handle lines of any length.
1552 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1553 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1556 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1557 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1558 no longer happens in comment lines. A list of replacements is now placed
1559 at the head of all of the source files, except those whose only change is
1560 to replace PERL_COMMAND in the very first #! line.
1562 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1563 messages on the queue, with a bottom-up merge sort, using code contributed
1564 by Michael Haardt. This should make operations like -bp somewhat faster on
1565 large queues. It won't affect queue runners, except when queue_run_in_order
1568 51. Installed eximstats 1.31 in the distribution.
1570 52. Added support for SRV lookups to the dnslookup router.
1572 53. If an ACL referred to $message_body or $message_body_end, the value was not
1573 reset for any messages that followed in the same SMTP session.
1575 54. The store-handling optimization for building very long strings was not
1576 differentiating between the different store pools. I don't think this
1577 actually made any difference in practice, but I've tidied it.
1579 55. While running the routers to verify a sender address, $sender_address
1580 was still set to the sender address. This is wrong, because when routing to
1581 send a bounce to the sender, it would be empty. Therefore, I have changed
1582 it so that, while verifying a sender address, $sender_address is set to <>.
1583 (There is no change to what happens when verifying a recipient address.)
1585 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1586 target A or AAAA records (if not already returned) without resetting the
1587 qualify_single or search_parents options of the DNS resolver. These are
1588 inappropriate in this case because the targets of MX and SRV records must
1589 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1590 target that, when qualified, matched something in the local domain. These
1591 two options are now turned off when doing these lookups.
1593 57. It seems that at least some releases of Reiserfs (which does not have the
1594 concept of a fixed number of inodes) returns zero and not -1 for the
1595 number of available inodes. This interacted badly with check_spool_inodes,
1596 which assumed that -1 was the "no such thing" setting. What I have done is
1597 to check that the total number of inodes is greater than zero before doing
1598 the test of how many are available.
1600 58. When a "warn" ACL statement has a log_message modifier, the message is
1601 remembered, and not repeated. This is to avoid a lot of repetition when a
1602 message has many recipients that cause the same warning to be written.
1603 Howewer, Exim was preserving the list of already written lines for an
1604 entire SMTP session, which doesn't seem right. The memory is now reset if a
1605 new message is started.
1607 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1608 debugging output unless log_rewrite was also set.
1610 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1611 of (int)(handle) when we know that handle contains (void *)(-1).
1613 61. The Exim daemon panic-logs an error return when it closes the incoming
1614 connection. However "connection reset by peer" seems to be common, and
1615 isn't really an error worthy of noting specially, so that particular error
1618 62. When Exim is trying to find all the local interfaces, it used to panic and
1619 die if the ioctl to get the interface flags failed. However, it seems that
1620 on at least one OS (Solaris 9) it is possible to have an interface that is
1621 included in the list of interfaces, but for which you get a failure error
1622 for this call. This happens when the interface is not "plumbed" into a
1623 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1624 failure of the "get flags" call assumes that the interface is down.
1626 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1627 makes life easier for people who are doing arithmetic on fields extracted
1628 from dates, where you often get leading zeros that should not be
1629 interpreted as octal.
1631 64. Added qualify_domain to the redirect router, to override the global
1634 65. If a pathologically long header line contained very many addresses (the
1635 report of this problem mentioned 10 000) and each of them was rewritten,
1636 Exim could use up a very large amount of memory. (It kept on making new
1637 copies of the header line as it rewrote, and never released the old ones.)
1638 At the expense of a bit more processing, the header rewriting function has
1639 been changed so that it no longer eats memory in this way.
1641 66. The generation of the Received: header has been moved from the time that a
1642 message starts to be received, to the time that it finishes. The timestamp
1643 in the Received: header should now be very close to that of the <= log
1644 line. There are two side-effects of this change:
1646 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1647 logged header lines no longer include the local Received: line, because
1648 it has not yet been created. The same applies to a copy of the message
1649 that is returned to a non-SMTP sender when a message is rejected.
1651 (b) When a filter file is tested using -bf, no additional Received: header
1652 is added to the test message. After some thought, I decided that this
1655 This change does not affect the value of $received_for. It is still set
1656 after address rewriting, but before local_scan() is called.
1658 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1660 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1661 gave an unhelpful panic error message, and a defer error. I have managed to
1662 change this behaviour so that it now rejects any supplied certificate,
1663 which seems right, as the list of acceptable certificates is empty.
1665 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1666 gave an unhelpful defer error. I have not managed to make this reject any
1667 supplied certificates, but the error message it gives is "no certificate
1668 supplied", which is not helpful.
1670 70. exigrep's output now also includes lines that are not associated with any
1671 message, but which match the given pattern. Implemented by a patch from
1672 Martin Sluka, which also tidied up the Perl a bit.
1674 71. Recipient callout verification, like sender verification, was using <> in
1675 the MAIL FROM command. This isn't really the right thing, since the actual
1676 sender may affect whether the remote host accepts the recipient or not. I
1677 have changed it to use the actual sender in the callout; this means that
1678 the cache record is now keyed on a recipient/sender pair, not just the
1679 recipient address. There doesn't seem to be a real danger of callout loops,
1680 since a callout by the remote host to check the sender would use <>.
1681 [SEE ABOVE: changed after hitting problems.]
1683 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1684 temporary errors. However, in the case of such a code being given after
1685 the end of a data transmission (i.e. after ".") Exim was failing to write
1686 a retry record for the message. (Yes, there was some broken host that was
1687 actually sending 8xx at this point.)
1689 73. An unknown lookup type in a host list could cause Exim to panic-die when
1690 the list was checked. (An example that provoked this was putting <; in the
1691 middle of a list instead of at the start.) If this happened during a DATA
1692 ACL check, a -D file could be left lying around. This kind of configuration
1693 error no longer causes Exim to die; instead it causes a defer errror. The
1694 incident is still logged to the main and panic logs.
1696 74. Buglet left over from Exim 3 conversion. The message "too many messages
1697 in one connection" was written to the rejectlog but not the mainlog, except
1698 when address rewriting (yes!) was being logged.
1700 75. Added write_rejectlog option.
1702 76. When a system filter was run not as root (that is, when system_filter_user
1703 was set), the values of the $n variables were not being returned to the
1704 main process; thus, they were not subsequently available in the $sn
1707 77. Added +return_path_on_delivery log selector.
1709 78. A connection timeout was being treated differently from recipients deferred
1710 when testing hosts_max_try with a message that was older than the host's
1711 retry timeout. (The host should not be counted, thus allowing all hosts to
1712 be tried at least once before bouncing.) This may have been the cause of an
1713 occasionally reported bug whereby a message would remain on the queue
1714 longer than the retry timeout, but would be bounced if a delivery was
1715 forced. I say "may" because I never totally pinned down the problem;
1716 setting up timeout/retry tests is difficult. See also the next item.
1718 79. The ultimate address timeout was not being applied to errors that involved
1719 a combination of host plus message (for example, a timeout on a MAIL
1720 command). When an address resolved to a number of possible hosts, and they
1721 were not all tried for each delivery (e.g. because of hosts_max_try), a
1722 message could remain on the queue longer than the retry timeout.
1724 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1727 81. Fixed an obscure SMTP outgoing bug which required at least the following
1728 conditions: (a) there was another message waiting for the same server;
1729 (b) the server returned 5xx to all RCPT commands in the first message so
1730 that the message was not completed; (c) the server dropped the connection
1731 or gave a negative response to the RSET that Exim sends to abort the
1732 transaction. The observed case was a dropped connection after DATA that had
1733 been sent in pipelining mode. That is, the server had advertised PIPELINING
1734 but was not implementing it correctly. The effect of the bug was incorrect
1735 behaviour, such as trying another host, and this could lead to a crash.
1741 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1742 and daemon.c were passed as pointers to ints; they should have been
1743 pointers to socklen_t variables (which are typically unsigned ints).
1745 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1748 3. Fixed a really odd bug that affected only the testing scheme; patching a
1749 certain fixed string in the binary changed the value of another string that
1750 happened to be identical to the end of the original first string.
1752 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1753 name", it returns that address as the IP address. On some operating
1754 systems (e.g. Solaris), it also passes back the IP address string as the
1755 "host name". However, on others (e.g. Linux), it passes back an empty
1756 string. Exim wasn't checking for this, and was changing the host name to an
1757 empty string, assuming it had been canonicized.
1759 5. Although rare, it is permitted to have more than one PTR record for a given
1760 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1761 all the names associated with an address, because they do in Solaris.
1762 However, it seems that they do not in Linux for data that comes from the
1763 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1764 I found this out when I moved to a new Linux workstation and tried to run
1765 the Exim test suite.
1767 To get round this problem I have changed the code so that it now does its
1768 own call to the DNS to look up PTR records when searching for a host name.
1769 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1770 addresses that are only in /etc/hosts are still found.
1772 This behaviour is, however, controlled by an option called host_lookup_
1773 order, which defaults to "bydns:byaddr". If people want to use the other
1774 order, or indeed, just use one or the other means of lookup, they can
1775 specify it in this variable.
1777 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1778 some operating systems, this comes back from gethostbyaddr() as an empty
1779 string, and this is what Exim used to test for. However, it seems that in
1780 other systems, "." is yielded. Exim now tests for this case too.
1782 7. The values of check_spool_space and check_log_space are now held internally
1783 as a number of kilobytes instead of an absolute number of bytes. If a
1784 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1785 kilobyte. This means that much larger values can be stored.
1787 8. Exim monitor: an attempt to get the action menu when not actually pointing
1788 at a message produces an empty menu entitled "No message selected". This
1789 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1790 no entries in it ("Shell widget menu has zero width and/or height"). So I
1791 have added a single, blank menu entry in this case.
1793 9. Added ${quote_local_part.
1795 10. MIME decoding is now applied to the contents of Subject: header lines when
1798 11. Now that a reference to $sender_host_address automatically causes a reverse
1799 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1800 host lookup before query-style lookups in lists that might use this
1801 variable. This has therefore been abolished, and the "net-" prefix is no
1802 longer necessary for query-style lookups.
1804 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1805 have been a typo for LFLAGS, so it has been changed.
1807 13. The install script calls Exim with "-C /dev/null" in order to find the
1808 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1809 to be output. Howeve, since Exim outputs its version number before the
1810 error, it didn't break the script. It just looked ugly. I fixed this by
1811 always allowing "-C /dev/null" if the caller is root.
1813 14. Ignore overlarge ACL variable number when reading spool file - insurance
1814 against a later release with more variables having written the file.
1816 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1817 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1818 this, as well as the form without the "IPv6" on the front.
1820 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1821 OS/Makefile-Darwin file.
1823 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1824 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1826 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1827 variable. However, some OS use ints for inodes. I've added cast to long int
1828 to get rid of the compiler warning.
1830 19. I had forgotten to lock out "/../" in configuration file names when
1831 ALT_CONFIG_PREFIX was set.
1833 20. Routers used for verification do not need to specify transports. However,
1834 if such a router generated a host list, and callout was configured, Exim
1835 crashed, because it could not find a port number from the (non-existent)
1836 transport. It now assumes port 25 in this circumstance.
1838 21. Added the -t option to exigrep.
1840 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1841 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1842 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1843 them is set, LOOKUP_LSEARCH is forced.
1845 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1847 24. Added sender and host information to the "rejected by local_scan()" log
1848 line; previously there was no indication of these.
1850 25. Added .include_if_exists.
1852 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1853 Linux. It turns out that not all file systems support this. Apparently some
1854 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1855 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1856 not supported on the file descriptor, is now ignored when Exim is trying to
1857 sync a directory. This applies only to Linux.
1859 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1861 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1862 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1865 29. Added support for maildirsize files from supplied patch (modified a bit).
1867 30. The use of :fail: followed by an empty string could lead Exim to respond to
1868 sender verification failures with (e.g.):
1870 550 Verification failed for <xxx>
1871 550 Sender verify failed
1873 where the first response line was missing the '-' that indicates it is not
1874 the final line of the response.
1876 31. The loop for finding the name of the user that called Exim had a hardwired
1877 limit of 10; it now uses the value of finduser_retries, which is used for
1878 all other user lookups.
1880 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1883 33. Exim was neglecting to zero errno before one call of strtol() when
1884 expanding a string and expecting an integer value. On some systems this
1885 resulted in spurious "integer overflow" errors. Also, it was casting the
1886 result into an int without checking.
1888 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1889 did not work. The code looks as if it has *never* worked, though it appears
1890 to have been documented since at least releast 1.62. I have made it work.
1892 35. The "timeout_DNS" error in retry rules, also documented since at least
1893 1.62, also never worked. As it isn't clear exactly what this means, and
1894 clearly it isn't a major issue, I have abolished the feature by treating it
1895 as "timeout", and writing a warning to the main and panic logs.
1897 36. The display of retry rules for -brt wasn't always showing the error code
1900 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1901 timeout_connect_A, timeout_connect_MX.
1903 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1904 to the empty sender.
1906 39. The daemon was not analysing the content of -oX till after it had closed
1907 stderr and disconnected from the controlling terminal. This meant that any
1908 syntax errors were only noted on the panic log, and the return code from
1909 the command was 0. By re-arranging the code a little, I've made the
1910 decoding happen first, so such errors now appear on stderr, and the return
1911 code is 1. However, the actual setting up of the sockets still happens in
1912 the disconnected process, so errors there are still only recorded on the
1915 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1916 connections (as happens on some IP stacks) was logged at start up time as
1917 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1918 it from "IPv6 and IPv4", which means that two separate sockets are being
1921 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1922 says whether AF_INET or AF_INET6 was passed as an argument.
1924 42. Exiwhat output was messed up when time zones were included in log
1927 43. Exiwhat now gives more information about the daemon's listening ports,
1928 and whether -tls-on-connect was used.
1930 44. The "port" option of the smtp transport is now expanded.
1932 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1933 silently ignored. Now an error message is written to the main and panic
1936 46. There's a new ACL modifier called "logwrite" which writes to a log file
1937 as soon as it is encountered.
1939 47. Added $local_user_uid and $local_user_gid at routing time.
1941 48. Exim crashed when trying to verify a sender address that was being
1944 49. Exim was recognizing only a space character after ".include". It now also
1945 recognizes a tab character.
1947 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1948 extracting the relevant information from the specification. The man page no
1949 longer contains scrambled data for the -d option, and I've added a section
1950 at the front about calling Exim under different names.
1952 51. Added "extra_headers" argument to the "mail" command in filter files.
1954 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1957 53. Installed eximstats 1.29.
1959 54. Added transport_filter_timeout as a generic transport option.
1961 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1962 Cc: header lines. This was required by RFC 822, but it not required by RFC
1965 56. Exim used to add From:, Date:, and Message-Id: header lines to any
1966 incoming messages that did not have them. Now it does so only if the
1967 message originates locally, that is, if there is no associated remote host
1968 address. When Resent- header lines are present, this applies to the Resent-
1969 lines rather than the non-Resent- lines.
1971 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
1972 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
1974 58. Messages for configuration errors now include the name of the main
1975 configuration file - useful now that there may be more than one file in a
1976 list (.included file names were always shown).
1978 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1979 for those rare installations that do not start the daemon as root or run it
1980 setuid root. I've cut out the call to initgroups() if the daemon is not
1983 60. The Exim user and group can now be bound into the binary as text strings
1984 that are looked up at the start of Exim's processing.
1986 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
1988 62. Added $mailstore_basename variable.
1990 63. Installed patch to sieve.c from Michael Haardt.
1992 64. When Exim failed to open the panic log after failing to open the main log,
1993 the original message it was trying to log was written to stderr and debug
1994 output, but if they were not available (the usual case in production), it
1995 was lost. Now it is written to syslog before the two lines that record the
1996 failures to open the logs.
1998 65. Users' Exim filters run in subprocesses under the user's uid. It is
1999 possible for a "deliver" command or an alias in a "personal" command to
2000 provoke an address rewrite. If logging of address rewriting is configured,
2001 this fails because the process is not running as root or exim. There may be
2002 a better way of dealing with this, but for the moment (because 4.30 needs
2003 to be released), I have disabled address rewrite logging when running a
2004 filter in a non-root, non-exim process.
2010 1. The buildconfig auxiliary program wasn't quoting the value set for
2011 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2012 not defined. This bug was masked in 4.22 by the effect that was fixed in
2015 2. Some messages that were rejected after a message id was allocated were
2016 shown as "incomplete" by exigrep. It no longer does this for messages that
2017 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2019 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2020 have allow_domain_literals set, the ID did not get logged in the <= line.
2021 Domain literals are now always recognized in Message-ID: header lines.
2023 4. The first argument for a ${extract expansion item is the key name or field
2024 number. Leading and trailing spaces in this item were not being ignored,
2025 causing some misleading effects.
2027 5. When deliver_drop_privilege was set, single queue runner processes started
2028 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2029 same command in the process it spins off) were not dropping privilege.
2031 6. When the daemon running as "exim" started a queue runner, it always
2032 re-executed Exim in the spun-off process. This is a waste of effort when
2033 deliver_drop_privilege is set. The new process now just calls the
2034 queue-runner function directly.
2040 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2043 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2044 ignored. Though the use of -oP was forcing the writing of a pid file, it
2045 was always written to the default place.
2047 3. If the message "no IP address found for host xxxx" is generated during
2048 incoming verification, it is now followed by identification of the incoming
2049 connection (so you can more easily find what provoked it).
2051 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2053 5. Added some features to "harden" Exim a bit more against certain attacks:
2055 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2056 be put in Local/Makefile. This is like the never_users runtime option,
2057 but it cannot be overridden. The default setting is "root".
2059 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2060 prefix string with which any file named in a -C command line option
2063 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2064 is retained for -C and -D only if the caller of Exim is root. Without
2065 it, the exim user may also use -C and -D and retain privilege.
2067 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2068 command line option is disabled.
2070 6. Macro names set by the -D option must start with an upper case letter, just
2071 like macro names defined in the configuration file.
2073 7. Added "dereference=" facility to LDAP.
2075 8. Two instances of the typo "uknown" in the source files are fixed.
2077 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2078 the Configure-Makefile script screwed up while processing it.
2080 10. Incorporated PCRE 4.4.
2082 11. The SMTP synchronization check was not operating right at the start of an
2083 SMTP session. For example, it could not catch a HELO sent before the client
2084 waited for the greeting. There is now a check for outstanding input at the
2085 point when the greeting is written. Because of the duplex, asynchronous
2086 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2087 way, but not yet received, when the check is performed.
2089 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2090 on TCP/IP sockets, because this apparently causes some broken clients to
2093 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2094 unchanged) from the Cygwin maintainer.
2096 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2097 was supported instead of testing for mbx. Effectively a typo.
2099 15. The spa authenticator server code was not checking that the input it
2100 received was valid base64.
2102 16. The debug output line for the "set" modifier in ACLs was not showing the
2103 name of the variable that was being set.
2105 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2107 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2108 lookup on its own. Something else was needed to trigger the lookup. For
2109 example, a match in host_lookup or the need for a host name in a host list.
2110 Now, if $sender_host_name is referenced and the host name has not yet been
2111 looked up, a lookup is performed. If the lookup fails, the variable remains
2112 empty, and $host_lookup_failed is set to "1".
2114 19. Added "eqi" as a case-independent comparison operator.
2116 20. The saslauthd authentication condition could segfault if neither service
2117 nor realm was specified.
2119 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2120 error message that was logged was misleading, and incoming SMTP
2121 connections were dropped. The message is now more accurate, and temporary
2122 errors are given to SMTP connections.
2124 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2125 (except RSET) with a 421 error, until QUIT is received. However, it was
2126 failing to send a response to QUIT.
2128 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2129 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2130 ACL is now run after the helo_try_verify_hosts code.
2132 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2133 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2134 case-independent, so other case variants are also recognized). Apparently
2135 some systems use these upper case variants.
2137 25. If more than two messages were waiting for the same host, and a transport
2138 filter was specified for the transport, Exim sent two messages over the
2139 same TCP/IP connection, and then failed with "socket operation on non-
2140 socket" when it tried to send the third.
2142 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2144 27. The extern definition of crypt16() in expand.c was not being excluded when
2145 the OS had its own crypt16() function.
2147 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2148 as a preferred synonym for return_size_limit, both as an option and as an
2151 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2153 30. Changed the default configuration ACL to relax the local part checking rule
2154 for addresses that are not in any local domains. For these addresses,
2155 slashes and pipe symbols are allowed within local parts, but the sequence
2156 /../ is explicitly forbidden.
2158 31. SPA server authentication was not clearing the challenge buffer before
2161 32. log_message in a "warn" ACL statement was writing to the reject log as
2162 well as to the main log, which contradicts the documentation and doesn't
2163 seem right (because no rejection is happening). So I have stopped it.
2165 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2166 However, I am unable to do any testing of this.
2168 34. Fixed an infelicity in the appendfile transport. When checking directories
2169 for a mailbox, to see if any needed to be created, it was accidentally
2170 using path names with one or more superfluous leading slashes; tracing
2171 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2173 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2174 contents are added to the log line that is written for every discarded
2175 recipient. (Previously a log_message setting was ignored.)
2177 36. The ${quote: operator now quotes the string if it is empty.
2179 37. The install script runs exim in order to find its version number. If for
2180 some reason other than non-existence or emptiness, which it checks, it
2181 could not run './exim', it was installing it with an empty version number,
2182 i.e. as "exim-". This error state is now caught, and the installation is
2185 38. An argument was missing from the function that creates an error message
2186 when Exim fails to connect to the socket for saslauthd authentication.
2187 This could cause Exim to crash, or give a corrupted message.
2189 39. Added isip, isip4, and isip6 to ${if conditions.
2191 40. The ACL variables $acl_xx are now saved with the message, and can be
2192 accessed later in routers, transports, and filters.
2194 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2195 strings in the file are not string-expanded.
2197 42. If a MAIL command specified a SIZE value that was too large to fit into an
2198 int variable, the check against message_size_limit failed. Such values are
2199 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2200 day this will have to be increased, but I don't think I want to be around
2201 when emails are that large.
2208 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2209 iconv() is not standard in FreeBSD.
2211 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2212 IPv6 enabled. The observed symptom was a segmentation fault on return from
2213 the function os_common_find_running_interfaces() in src/os.c.
2215 3. In the check_special_case() function in daemon.c I had used "errno" as an
2216 argument name, which causes warnings on some systems. This was basically a
2217 typo, since it was named "eno" in the comments!
2219 4. The code that waits for the clock to tick (at a resolution of some fraction
2220 of a second) so as to ensure message-id uniqueness was always waiting for
2221 at least one whole tick, when it could have waited for less. [This is
2222 almost certainly not relevant at current processor speeds, where it is
2223 unlikely to ever wait at all. But we try to future-proof.]
2225 5. The function that sleeps for a time interval that includes fractions of a
2226 second contained a race. It did not block SIGALRM between setting the
2227 timer, and suspending (a couple of lines later). If the interval was short
2228 and the sigsuspend() was delayed until after it had expired, the suspension
2229 never ended. On busy systems this could lead to processes getting stuck for
2232 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2233 process, before it forks any delivery processes. The open lookup caching
2234 mechanism meant that the open file or database connection was passed into
2235 the delivery process. The problem was that delivery processes always tidy
2236 up cached lookup data. This could cause a problem for the next delivery
2237 process started by the queue runner, because the external queue runner
2238 process does not know about the closure. So the next delivery process
2239 still has data in the lookup cache. In the case of a file lookup, there was
2240 no problem because closing a file descriptor in a subprocess doesn't affect
2241 the parent. However, if the lookup was caching a connection to a database,
2242 the connection was closed, and the second delivery process was likely to
2243 see errors such as "PGSQL: query failed: server closed the connection
2244 unexpectedly". The problem has been fixed by closing all cached lookups
2245 in a queue runner before running a delivery process.
2247 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2248 seem to have the "const" qualifier which it has on other OS. I've
2251 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2252 *of the same type* (client or server) with the same public name that an
2253 error should be diagnosed.
2255 9. When Exim looked up a host name for an IP address, but failed to find the
2256 original IP address when looking up the host name (a safety check), it
2257 output the message "<ip address> does not match any IP for NULL", which was
2258 confusing, to say the least. The bug was that the host name should have
2259 appeared instead of "NULL".
2261 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2262 user that is built into the binary, and the -C or -D options is used, root
2263 privilege is dropped before the configuration file is read. In addition,
2264 logging is switched to stderr instead of the normal log files. If the
2265 configuration then re-defines the Exim user, the unprivileged environment
2266 is probably not what is expected, so Exim logs a panic warning message (but
2269 However, if deliver_drop_privilege is set, the unprivileged state may well
2270 be exactly what is intended, so the warning has been cut out in that case,
2271 and Exim is allowed to try to write to its normal log files.
2277 1. smtp_return_error_details was not giving details for temporary sender
2278 or receiver verification errors.
2280 2. Diagnose a configuration error if two authenticators have the same public
2283 3. Exim used not to create the message log file for a message until the first
2284 delivery attempt. This could be confusing when incoming messages were held
2285 for policy or load reasons. The message log file is now created at the time
2286 the message is received, and an initial "Received" line is written to it.
2288 4. The automatically generated man page for command line options had a minor
2289 bug that caused no ill effects; however, a more serious problem was that
2290 the procedure for building the man page automatically didn't always
2291 operate. Consequently, release 4.20 contains an out-of-date version. This
2292 shouldn't happen again.
2294 5. When building Exim with embedded Perl support, the script that builds the
2295 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2296 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2298 6. The freeze_tell option was not being used for messages that were frozen on
2299 arrival, either by an ACL or by local_scan().
2301 7. Added the smtp_incomplete_transaction log selector.
2303 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2304 was accepting AUTH without a new EHLO.
2306 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2307 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2310 10. Logging of TCP/IP connections (when configured) now happens in the main
2311 daemon process instead of the child process, so that the TCP/IP connection
2312 count is more accurate (but it can never be perfect).
2314 11. The use of "drop" in a nested ACL was not being handled correctly in the
2315 outer ACL. Now, if condition failure induced by the nested "drop" causes
2316 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2317 or "require"), the connection is dropped.
2319 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2320 that yield "discard" can now be used with an "accept" or a "discard" verb,
2321 but an error is generated for any others (because I can't see a useful way
2322 to define what should happen).
2324 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2325 are now processed in the same way as lines in the Exim configuration file.
2326 In particular, continuation lines are supported.
2328 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2330 15. Added -ti meaning -t -i.
2332 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2333 domains, and warn by logging if others are found.
2335 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2336 returned from GIFCONF when Exim is trying to find the list of interfaces on
2337 a host. The code in os.c has been modified to copy each ifreq to an aligned
2338 structure in all cases.
2340 Also, in some cases, the returned ifreq's were being copied to a 'struct
2341 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2342 means the last couple of bytes of an IPv6 address could be chopped if the
2343 ifreq contained only a normal sockaddr (14 bytes storage).
2345 18. Named domain lists were not supported in the hosts_treat_as_local option.
2346 An entry such as +xxxx was not recognized, and was treated as a literal
2349 19. Ensure that header lines added by a DATA ACL are included in the reject log
2350 if the ACL subsequently rejects the message.
2352 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2353 MD5 (which is deprecated).
2355 21. When testing a filter file using -bf, Exim was writing a message when it
2356 took the sender from a "From " line in the message, but it was not doing so
2357 when it took $return_path from a Return-Path: header line. It now does.
2359 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2360 with a valid header line field name (a series of printing characters
2361 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2363 23. Changed "disc" in the source to "disk" to conform to the documentation and
2364 the book and for uniformity.
2366 24. Ignore Sendmail's -Ooption=value command line item.
2368 25. When execve() failed while trying to run a command in a pipe transport,
2369 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2370 could be confused with a return value of 69 from the command itself. This
2371 has been changed to 127, the value the shell returns if it is asked to run
2372 a non-existent command. The wording for the related log line suggests a
2373 non-existent command as the problem.
2375 26. If received_header_text expands to an empty string, do not add a Received:
2376 header line to the message. (Well, it adds a token one on the spool, but
2377 marks it "old" so that it doesn't get used or transmitted.)
2379 27. Installed eximstats 1.28 (addition of -nt option).
2381 28. There was no check for failure on the call to getsockname() in the daemon
2382 code. This can fail if there is a shortage of resources on the system, with
2383 ENOMEM, for example. A temporary error is now given on failure.
2385 29. Contrary to the C standard, it seems that in some environments, the
2386 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2387 program. Exim now does this explicitly; it affects the formatting of
2388 timestamps using strftime().
2390 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2391 complaints. I've now initialized all the variables, to avoid this.
2393 32. Header lines added by a system filter were not being "seen" during
2394 transport-time rewrites.
2396 33. The info_callback() function passed to OpenSSL is set up with type void
2397 (*)(SSL *, int, int), as described somewhere. However, when calling the
2398 function (actually a macro) that sets it up, the type void(*)() is
2399 expected. I've put in a cast to prevent warnings from picky compilers.
2401 34. If a DNS black list lookup found a CNAME record, but there were no A
2402 records associated with the domain it pointed at, Exim crashed.
2404 35. If a DNS black list lookup returned more than one A record, Exim ignored
2405 all but the first. It now scans all returned addresses if a particular IP
2406 value is being sought. In this situation, the contents of the
2407 $dnslist_value variable are a list of all the addresses, separated by a
2410 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2411 to accept a host name and all its aliases if the forward lookup for any of
2412 them yielded the IP address of the incoming connection. Now it accepts only
2413 those names whose forward lookup yields the correct IP address. Any other
2414 names are discarded. This closes a loophole whereby a rogue DNS
2415 administrator could create reverse DNS records to break through a
2416 wildcarded host restriction in an ACL.
2418 37. If a user filter or a system filter that ran in a subprocess used any of
2419 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2420 the wrong values were passed to the pipe command ($thisaddress had the
2421 value of $0, $0 had the value of $1, etc). This bug was introduced by
2422 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2424 38. Improved the line breaking for long SMTP error messages from ACLs.
2425 Previously, if there was no break point between 40 and 75 characters, Exim
2426 left the rest of the message alone. Two changes have been made: (a) I've
2427 reduced the minimum length to 35 characters; (b) if it can't find a break
2428 point between 35 and 75 characters, it looks ahead and uses the first one
2429 that it finds. This may give the occasional overlong line, but at least the
2430 remaining text gets split now.
2432 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2433 file descriptors might be low, and that setting 1000 would always raise it.
2434 It turns out that in some environments, the limit is already over 1000 and
2435 that lowering it causes trouble. So now Exim takes care not to decrease it.
2437 40. When delivering a message, the value of $return_path is set to $sender_
2438 address at the start of routing (routers may change the value). By an
2439 oversight, this default was not being set up when an address was tested by
2440 -bt or -bv, which affected the outcome if any router or filter referred to
2443 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2444 log only when "message" or "log_message" are set. However, if one of the
2445 conditions was an address verification, or a call to a nested ACL, the
2446 messages generated by the underlying test were being passed through. This
2447 no longer happens. The underlying message is available in $acl_verify_
2448 message for both "message" and "log_message" expansions, so it can be
2449 passed through if needed.
2451 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2452 new expansion $bh_ to give the encoded byte string without charset
2453 translation. Translation happens only if iconv() is available; HAVE_ICONV
2454 indicates this at build time. HEADERS_CHARSET gives the charset to
2455 translate to; headers_charset can change it in the configuration, and
2456 "headers charset" can change it in an individual filter file.
2458 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2459 that creates RFC 2047 encoded "words" labels them as that charset instead
2460 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2461 expansion operator; (ii) when Exim creates a From: line for a local
2462 message; (iii) when a header line is rewritten to include a "phrase" part.
2464 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2465 buggy, causing it to skip the first lines of messages whose message ID
2466 ended in 'D'. This would not have bitten before Exim release 4.14, because
2467 message IDs were unlikely to end in 'D' before then. The effect was to have
2468 incorrect size information for certain domains.
2470 45. #include "config.h" was missing at the start of the crypt16.c module. This
2471 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2474 46. If there was a timeout during a "random" callout check, Exim treated it as
2475 a failure of the random address, and carried on sending RSET and the real
2476 address. If the delay was just some slowness somewhere, the response to the
2477 original RCPT would be taken as a response to RSET and so on, causing
2478 mayhem of various kinds.
2480 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2481 when I implemented it. It didn't allow for the fact that some option values
2482 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2483 the right test for positive values.
2485 48. Domain names in DNS records are case-independent. Exim always looks them up
2486 in lower case. Some resolvers return domain names in exactly the case they
2487 appear in the zone file, that is, they may contain uppercase letters. Not
2488 all resolvers do this - some return always lower case. Exim was treating a
2489 change of case by a resolver as a change of domain, similar to a widening
2490 of a domain abbreviation. This triggered its re-routing code and so it was
2491 trying to route what was effectively the same domain again. This normally
2492 caused routing to fail (because the router wouldn't handle the domain
2493 twice). Now Exim checks for this case specially, and just changes the
2494 casing of the domain that it ultimately uses when it transmits the message
2497 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2500 50. If a filter generated a file delivery with a non-absolute name (possible if
2501 no home directory exists for the router), the forbid_file option was not
2504 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2505 the existing equality matching.
2507 52. Exim was using ints instead of ino_t variables in some places where it was
2508 dealing with inode numbers.
2510 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2511 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2512 called TMPDIR, and if it finds it is different, it changes its value.
2514 54. The smtp_printf() function is now made available to local_scan() so
2515 additional output lines can be written before returning. There is also an
2516 smtp_fflush() function to enable the detection of a dropped connection.
2517 The variables smtp_input and smtp_batched_input are exported to
2520 55. Changed the default runtime configuration: the message "Unknown user"
2521 has been removed from the ACL, and instead placed on the localuser router,
2522 using the cannot_route_message feature. This means that any verification
2523 failures that generate their own messages won't get overridden. Similarly,
2524 the "Unrouteable address" message that was in the ACL for unverifiable
2525 relay addresses has also been removed.
2527 56. Added hosts_avoid_esmtp to the smtp transport.
2529 57. The exicyclog script was not checking for the esoteric option
2530 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2531 will work only if exicyclog is run under the appropriate euid.
2533 58. Following a discussion on the list, the rules by which Exim recognises line
2534 endings on incoming messages have been changed. The -dropcr and drop_cr
2535 options are now no-ops, retained only for backwards compatibility. The
2536 following line terminators are recognized: LF CRLF CR. However, special
2537 processing applies to CR:
2539 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2540 nor a local message in the state where . is a terminator.
2542 (ii) If a bare CR is encountered in a header line, an extra space is added
2543 after the line terminator so as not to end the header. The reasoning
2544 behind this is that bare CRs in header lines are most likely either
2545 to be mistakes, or people trying to play silly games.
2547 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2548 was being incorrectly given as 18 bytes larger than it should have been.
2549 This is a VOB (very old bug).
2551 60. This may never have affected anything current, but just in case it has:
2552 When the local host is found other than at the start of a list of hosts,
2553 the local host, those with the same MX, and any that follow, are discarded.
2554 When the list in question was part of a longer list of hosts, the following
2555 hosts (not currently being processed) were also being discarded. This no
2556 longer happens. I'm not sure if this situation could ever has previously
2559 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2562 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2563 header. This is something that is recommended in a new Internet Draft, and
2564 is something that is documented as being done by Sendmail. There are two
2565 possible values. For messages generated by the autoreply transport, Exim
2568 Auto-Submitted: auto-replied
2570 whereas for all other generated messages (e.g. bounces) it adds
2572 Auto-Submitted: auto-generated
2574 63. The "personal" condition in filters now includes a test for the
2575 Auto-Submitted: header. If it contains the string "auto-" the message it
2576 not considered personal.
2578 64. Added rcpt_include_affixes as a generic transport option.
2580 65. Added queue_only_override (default true).
2582 66. Added the syslog_duplication option.
2584 67. If what should have been the first header line of a message consisted of
2585 a space followed by a colon, Exim was mis-interpreting it as a header line.
2586 It isn't of course - it is syntactically invalid and should therefore be
2587 treated as the start of the message body. The misbehaviour could have
2588 caused a number of strange effects, including loss of data in subsequent
2589 header lines, and spool format errors.
2591 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2592 client host had authenticated. This control can now be exercised by an ACL
2593 for more flexibility.
2595 69. By default, callouts do not happen when testing with -bh. There is now a
2596 variant, -bhc, which does actually run the callout code, including
2597 consulting and updating the callout cache.
2599 70. Added support for saslauthd authentication, courtesy of Alexander
2602 71. If statvfs() failed on the spool or log directories while checking their
2603 size for availability, Exim confusingly gave the error "space shortage".
2604 Furthermore, in debugging mode it crashed with a floating point exception.
2605 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2606 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2607 serious problem, Exim now writes to the main and panic logs when this
2608 happens, with details of the failure. It then refuses to accept the
2609 incoming message, giving the message "spool directory problem" or "log
2610 directory problem" with a 421 code for SMTP messages.
2612 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2613 0, 1, and 2 exist, because some OS complain for execs without them (see
2614 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2615 descriptors. However, the code omitted to check that the open succeeded,
2616 causing mysterious errors if for some reason the permissions on /dev/null
2617 got screwed. Now Exim writes a message to the main and panic logs, and
2618 bombs out if it can't open /dev/null.
2620 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2621 interact so that it is all more flexible. It is supposed to remain
2622 backwards compatible. Also added extra_local_interfaces.
2624 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2625 to bomb out with an assertion failure - to the client this appears as a
2626 connection drop. This problem occurs in the part of the code that was taken
2627 from the Samba project. Fortunately, the assertion is in a very simple
2628 function, so I have fixed this by reproducing the function inline in the
2629 one place where it is called, and arranging for authentication to fail
2630 instead of killing the process with assert().
2632 75. The SPA client code was not working when the server requested OEM rather
2633 than Unicode encoding.
2635 76. Added code to make require_files with a specific uid setting more usable in
2636 the case where statting the file as root fails - usually a non-root-mounted
2637 NFS file system. When this happens and the failure is EACCES, Exim now
2638 forks a subprocess and does the per-uid checking as the relevant uid.
2640 77. Added process_log_path.
2642 78. If log_file_path was not explicitly set, a setting of check_log_space or
2643 check_log_inodes was ignored.
2645 79. If a space check for the spool or log partitions fails, the incident is now
2646 logged. Of course, in the latter case the data may get lost...
2648 80. Added the %p formatting code to string_format() so that it can be used to
2649 print addresses in debug_print(). Adjusted all the address printing in the
2650 debugging in store.c to use %p rather than %d.
2652 81. There was a concern that a line of code in smtp_in.c could overflow a
2653 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2654 initially expressed, the concern was not well-founded, because trailing
2655 spaces are removed early. However, if the trailing spaces were followed by
2656 a NULL, they did not get removed, so the overflow was possible. Two fixes
2659 (a) I re-wrote the offending code in a cleaner fashion.
2660 (b) If an incoming SMTP command contains a NULL character, it is rejected
2663 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2664 runs initgroups(), so that if the Exim user is in any additional groups,
2665 they will be used during message reception.
2671 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob