1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.99 2005/03/22 16:52:06 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
11 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
13 TK/02 Fix ACL "control" statment not being available in MIME ACL.
15 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
17 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
18 to test Sieve filters that use "vacation".
20 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
21 that changes the way the GnuTLS parameters are stored in the cache file.
22 The new format can be generated externally. For backward compatibility,
23 if the data in the cache doesn't make sense, Exim assumes it has read an
24 old-format file, and it generates new data and writes a new file. This
25 means that you can't go back to an older release without removing the
28 PH/03 A redirect router that has both "unseen" and "one_time" set does not
29 work if there are any delivery delays because "one_time" forces the
30 parent to be marked "delivered", so its unseen clone is never tried
31 again. For this reason, Exim now forbids the simultaneous setting of
34 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
35 redirected to themselves ("homonym" addresses). Read the long ChangeLog
36 entry if you want to know the details. The fix, however, neglected to
37 consider the case when local delivery batching is involved. The test for
38 "previously delivered" was not happening when checking to see if an
39 address could be batched with a previous (undelivered) one; under
40 certain circumstances this could lead to multiple deliveries to the same
41 address. A one-line patch to add the appropriate test fixes the bug.
43 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
44 in its include files, and this causes problems building Exim.
46 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
47 header_syntax) but Exim was just ignoring anything given after a slash.
48 In particular, this caused confusion with an attempt to use "verify =
49 reverse_host_lookup/defer_ok". An error is now given when options are
50 supplied for verify items that do not have them. (Maybe reverse_host_
51 lookup should have a defer_ok option, but that's a different point.)
53 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
54 defined by RFC 821) to 2048, because there were problems with some AUTH
55 commands, and RFC 1869 says the size should be increased for extended
56 SMTP commands that take arguments.
58 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
61 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
62 "unknown" error; now it says that the functionality isn't in the binary.
64 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
65 an address' error message when a string expansion fails (syntax or
66 whatever). Otherwise not only does the password appear in the log, it may
67 also be put in a bounce message.
69 PH/11 Installed exipick version 20050225.0 from John Jetmore.
71 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
72 of its addresses was ever tried. (Bugzilla bug #2.)
74 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
75 the result incorrectly in the debug output. (It correctly added a newline
76 to what was transported.)
78 TF/01 Added $received_time.
81 A note about Exim versions 4.44 and 4.50
82 ----------------------------------------
84 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
85 changes of various kinds. As a consequence, a big documentation update was
86 needed. This delayed the release for rather longer than seemed good, especially
87 in the light of a couple of (minor) security issues. Therefore, the changes
88 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
89 release. So 4.44 and 4.50 are in effect two different branches that both start
92 I have left the 4.50 change log unchanged; it contains all the changes since
93 4.43. The change log for 4.44 is below; many of its items are identical to
94 those for 4.50. This seems to be the most sensible way to preserve the
95 historical information.
101 1. Minor wording change to the doc/README.SIEVE file.
103 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
104 computation of the current number of files was incorrect.
106 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
107 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
110 4. Give more explanation in the error message when the command for a transport
111 filter fails to execute.
113 5. There are several places where Exim runs a non-Exim command in a
114 subprocess. The SIGUSR1 signal should be disabled for these processes. This
115 was being done only for the command run by the queryprogram router. It is
116 now done for all such subprocesses. The other cases are: ${run, transport
117 filters, and the commands run by the lmtp and pipe transports.
119 6. Added CONFIGURE_GROUP build-time option.
121 7. Some older OS have a limit of 256 on the maximum number of file
122 descriptors. Exim was using setrlimit() to set 1000 as a large value
123 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
124 systems. I've change it so that if it can't get 1000, it tries for 256.
126 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
127 was an oversight, and furthermore, ever since the addition of extra
128 controls (e.g. 4.43/32), the checks on when to allow different forms of
129 "control" were broken. There should now be diagnostics for all cases when a
130 control that does not make sense is encountered.
132 9. Added the /retain_sender option to "control=submission".
134 10. $recipients is now available in the predata ACL (oversight).
136 11. Tidy the search cache before the fork to do a delivery from a message
137 received from the command line. Otherwise the child will trigger a lookup
138 failure and thereby defer the delivery if it tries to use (for example) a
139 cached ldap connection that the parent has called unbind on.
141 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
142 of $address_data from the recipient verification was clobbered by the
145 13. The value of address_data from a sender verification is now available in
146 $sender_address_data in subsequent conditions in the ACL statement.
148 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
150 15. Added a new option "connect=<time>" to callout options, to set a different
153 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
154 was its contents. (It was OK if the option was not defined at all.)
156 17. A "Completed" log line is now written for messages that are removed from
157 the spool by the -Mrm option.
159 18. New variables $sender_verify_failure and $recipient_verify_failure contain
160 information about exactly what failed.
162 19. Added -dd to debug only the daemon process.
164 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
165 handles timeouts, both on the server side and network timeouts. Renamed the
166 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
168 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
169 It is now set to "smtps".
171 22. $host_address is now set to the target address during the checking of
174 23. When checking ignore_target_hosts for an ipliteral router, no host name was
175 being passed; this would have caused $sender_host_name to have been used if
176 matching the list had actually called for a host name (not very likely,
177 since this list is usually IP addresses). A host name is now passed as
180 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
181 code that specifies a non-restarting handler (typically sigaction() in
182 modern systems) in an attempt to fix a rare and obscure crash bug.
184 25. Narrowed the window for a race in the daemon that could cause it to ignore
185 SIGCHLD signals. This is not a major problem, because they are used only to
186 wake it up if nothing else does.
188 26. A malformed maildirsize file could cause Exim to calculate negative values
189 for the mailbox size or file count. Odd effects could occur as a result.
190 The maildirsize information is now recalculated if the size or filecount
193 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
194 support for a long time. Removed HAVE_SYS_VFS_H.
196 28. Installed the latest version of exipick from John Jetmore.
198 29. In an address list, if the pattern was not a regular expression, an empty
199 subject address (from a bounce message) matched only if the pattern was an
200 empty string. Non-empty patterns were not even tested. This was the wrong
201 because it is perfectly reasonable to use an empty address as part of a
202 database query. An empty address is now tested by patterns that are
203 lookups. However, all the other forms of pattern expect the subject to
204 contain a local part and a domain, and therefore, for them, an empty
205 address still always fails if the pattern is not itself empty.
207 30. Exim went into a mad DNS loop when attempting to do a callout where the
208 host was specified on an smtp transport, and looking it up yielded more
211 31. Re-factored the code for checking spool and log partition space into a
212 function that finds that data and another that does the check. The former
213 is then used to implement four new variables: $spool_space, $log_space,
214 $spool_inodes, and $log_inodes.
216 32. The RFC2047 encoding function was originally intended for short strings
217 such as real names; it was not keeping to the 75-character limit for
218 encoded words that the RFC imposes. It now respects the limit, and
219 generates multiple encoded words if necessary. To be on the safe side, I
220 have increased the buffer size for the ${rfc2047: expansion operator from
223 33. It is now permitted to omit both strings after an "if" condition; if the
224 condition is true, the result is "true". As before, when the second string
225 is omitted, a false condition yields an empty string. This makes it less
226 cumbersome to write custom ACL and router conditions.
228 34. Failure to deliver a bounce message always caused it to be frozen, even if
229 there was an errors_to setting on the router. The errors_to setting is now
232 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
233 canonical form (fully expanded) before being placed in
234 $sender_host_address.
236 36. The table in the code that translates DNS record types into text (T_A to
237 "A" for instance) was missing entries for NS and CNAME. It is just possible
238 that this could have caused confusion if both these types were looked up
239 for the same domain, because the text type is used as part of Exim's
240 per-process caching. But the chance of anyone hitting this buglet seems
243 37. The dnsdb lookup has been extended in a number of ways.
245 (1) There is a new type, "zns", which walks up the domain tree until it
246 finds some nameserver records. It should be used with care.
248 (2) There is a new type, "mxh", which is like "mx" except that it returns
249 just the host names, not the priorities.
251 (3) It is now possible to give a list of domains (or IP addresses) to be
252 looked up. The behaviour when one of the lookups defers can be
253 controlled by a keyword.
255 (4) It is now possible to specify the separator character for use when
256 multiple records are returned.
258 38. The dnslists ACL condition has been extended: it is now possible to supply
259 a list of IP addresses and/or domains to be looked up in a particular DNS
262 39. Added log_selector=+queue_time_overall.
264 40. When running the queue in the test harness, wait just a tad after forking a
265 delivery process, to get repeatability of debugging output.
267 41. Include certificate and key file names in error message when GnuTLS fails
268 to set them up, because the GnuTLS error message doesn't include the name
269 of the failing file when there is a problem reading it.
271 42. Allow both -bf and -bF in the same test run.
273 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
275 44. The "Exiscan patch" is now merged into the mainline Exim source.
277 45. Sometimes the final signoff response after QUIT could fail to get
278 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
279 < 0 before doing a fflush(). This bug looks as though it goes back to the
280 introduction of TLS in release 3.20, but "sometimes" must have been rare
281 because the tests only now provoked it.
283 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
284 (this can affect the format of dates).
286 47. exim_tidydb, when checking for the continued existence of a message for
287 which it has found a message-specific retry record, was not finding
288 messages that were in split spool directories. Consequently, it was
289 deleting retry records that should have stayed in existence.
291 48. Steve fixed some bugs in eximstats.
293 49. The SPA authentication driver was not abandoning authentication and moving
294 on to the next authenticator when an expansion was forced to fail,
295 contradicting the general specification for all authenticators. Instead it
296 was generating a temporary error. It now behaves as specified.
298 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
299 (the order specifies the preference for clients). The order is now AES256,
300 AES128, 3DES, ARCFOUR128.
302 51. Small patch to Sieve code - explicitly set From: when generating an
305 52. Exim crashed if a remote delivery caused a very long error message to be
306 recorded - for instance if somebody sent an entire SpamAssassin report back
307 as a large number of 550 error lines. This bug was coincidentally fixed by
308 increasing the size of one of Exim's internal buffers (big_buffer) that
309 happened as part of the Exiscan merge. However, to be on the safe side, I
310 have made the code more robust (and fixed the comments that describe what
313 53. Now that there can be additional text after "Completed" in log lines (if
314 the queue_time_overall log selector is set), a one-byte patch to exigrep
315 was needed to allow it to recognize "Completed" as not the last thing in
318 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
319 patch that reportedly fixes this has been added. I am not expert enough to
320 create a test for it. This is what the patch creator wrote:
322 "I found a little strange behaviour of ldap code when working with
323 Windows 2003 AD Domain, where users was placed in more than one
324 Organization Units. When I tried to give exim partial DN, the exit code
325 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
326 But simultaneously result of request was absolutely normal ldap result,
327 so I produce this patch..."
329 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
330 REFERENCE, so I have modified the code to exclude the patch when that macro
333 55. Some experimental protocols are using DNS PTR records for new purposes. The
334 keys for these records are domain names, not reversed IP addresses. The
335 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
336 leaves it alone. Component reversal etc. now happens only for IP addresses.
339 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
341 57. Double the size of the debug message buffer (to 2048) so that more of very
342 long debug lines gets shown.
344 58. The exicyclog utility now does better if the number of log files to keep
345 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
347 59. Two changes related to the smtp_active_hostname option:
349 (1) $smtp_active_hostname is now available as a variable.
350 (2) The default for smtp_banner uses $smtp_active_hostname instead
351 of $primary_hostname.
353 60. The host_aton() function is supposed to be passed a string that is known
354 to be a valid IP address. However, in the case of IPv6 addresses, it was
355 not checking this. This is a hostage to fortune. Exim now panics and dies
356 if the condition is not met. A case was found where this could be provoked
357 from a dnsdb PTR lookup with an IPv6 address that had more than 8
358 components; fortuitously, this particular loophole had already been fixed
359 by change 4.50/55 above.
361 If there are any other similar loopholes, the new check in host_aton()
362 itself should stop them being exploited. The report I received stated that
363 data on the command line could provoke the exploit when Exim was running as
364 exim, but did not say which command line option was involved. All I could
365 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
369 61. There was a buffer overflow vulnerability in the SPA authentication code
370 (which came originally from the Samba project). I have added a test to the
371 spa_base64_to_bits() function which I hope fixes it.
374 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
375 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
377 63. The daemon start-up calls getloadavg() while still root for those OS that
378 need the first call to be done as root, but it missed one case: when
379 deliver_queue_load_max is set with deliver_drop_privilege. This is
380 necessary for the benefit of the queue runner, because there is no re-exec
381 when deliver_drop_privilege is set.
383 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
386 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
387 was in use, was not putting the data itself into the right store pool;
388 consequently, it could be overwritten for a subsequent message in the same
389 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
392 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
394 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
395 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
396 the function were treating the return as a boolean value, which happened to
397 work because 0=false and not-0=true, but is not correct code.
399 68. The host_aton() function was not handling scoped IPv6 addresses (those
400 with, for example, "%eth0" on the end) correctly.
402 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
403 negated items (that is, ~something) in unsigned ints. Some compilers
404 apparently mutter when there is no cast.
406 70. If an address verification called from an ACL failed, and did not produce a
407 user-specific message (i.e. there was only a "system" message), nothing was
408 put in $acl_verify_message. In this situation, it now puts the system
411 71. Change 4.23/11 added synchronization checking at the start of an SMTP
412 session; change 4.31/43 added the unwanted input to the log line - except
413 that it did not do this in the start of session case. It now does.
415 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
416 This is wrong and can cause the other end to generate a synchronization
417 error if it is another Exim or anything else that does the synchronization
418 check. A QUIT command is no longer sent after a timeout.
420 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
423 74. The defer_ok option of callout verification was not working if it was used
424 when verifying addresses in header lines, that is, for this case:
426 verify = header_sender/callout=defer_ok
428 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
429 those file descriptors could be used for SMTP connections. If anything
430 wrote to stderr (the example that came up was "warn" in embedded Perl), it
431 could be sent to the SMTP client, causing chaos. The daemon now opens
432 stdin, stdout, and stderr to /dev/null when it puts itself into the
435 76. Arrange for output from Perl's "warn" command to be written to Exim's main
436 log by default. The user can override this with suitable Perl magic.
438 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
439 the log message when discard triggers, was not working for the DATA ACL or
440 for the non-SMTP ACL.
442 78. Error message wording change in sieve.c.
444 79. If smtp_accept_max_per_host was set, the number of connections could be
445 restricted to fewer than expected, because the daemon was trying to set up
446 a new connection before checking whether the processes handling previous
447 connections had finished. The check for completed processes is now done
448 earlier. On busy systems, this bug wouldn't be noticed because something
449 else would have woken the daemon, and it would have reaped the completed
452 80. If a message was submitted locally by a user whose login name contained one
453 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
454 It caused a spool format error. I have fixed the spool reading code. A
455 related problem was that the "from" clause in the Received: line became
456 illegal because of the space(s). It is now covered by ${quote_local_part.
458 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
461 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
463 83. Patch from Sieve maintainer to fix unterminated string problem in
466 84. Some minor changes to the Linux configuration files to help with other
467 OS variants using glibc.
469 85. One more patch for Sieve to update vacation handling to latest spec.
472 ----------------------------------------------------
473 See the note above about the 4.44 and 4.50 releases.
474 ----------------------------------------------------
480 1. Change 4.43/35 introduced a bug that caused file counts to be
481 incorrectly computed when quota_filecount was set in an appendfile
484 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
485 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
488 3. Give more explanation in the error message when the command for a transport
489 filter fails to execute.
491 4. There are several places where Exim runs a non-Exim command in a
492 subprocess. The SIGUSR1 signal should be disabled for these processes. This
493 was being done only for the command run by the queryprogram router. It is
494 now done for all such subprocesses. The other cases are: ${run, transport
495 filters, and the commands run by the lmtp and pipe transports.
497 5. Some older OS have a limit of 256 on the maximum number of file
498 descriptors. Exim was using setrlimit() to set 1000 as a large value
499 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
500 systems. I've change it so that if it can't get 1000, it tries for 256.
502 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
503 was an oversight, and furthermore, ever since the addition of extra
504 controls (e.g. 4.43/32), the checks on when to allow different forms of
505 "control" were broken. There should now be diagnostics for all cases when a
506 control that does not make sense is encountered.
508 7. $recipients is now available in the predata ACL (oversight).
510 8. Tidy the search cache before the fork to do a delivery from a message
511 received from the command line. Otherwise the child will trigger a lookup
512 failure and thereby defer the delivery if it tries to use (for example) a
513 cached ldap connection that the parent has called unbind on.
515 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
516 of $address_data from the recipient verification was clobbered by the
519 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
520 was its contents. (It was OK if the option was not defined at all.)
522 11. A "Completed" log line is now written for messages that are removed from
523 the spool by the -Mrm option.
525 12. $host_address is now set to the target address during the checking of
528 13. When checking ignore_target_hosts for an ipliteral router, no host name was
529 being passed; this would have caused $sender_host_name to have been used if
530 matching the list had actually called for a host name (not very likely,
531 since this list is usually IP addresses). A host name is now passed as
534 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
535 code that specifies a non-restarting handler (typically sigaction() in
536 modern systems) in an attempt to fix a rare and obscure crash bug.
538 15. Narrowed the window for a race in the daemon that could cause it to ignore
539 SIGCHLD signals. This is not a major problem, because they are used only to
540 wake it up if nothing else does.
542 16. A malformed maildirsize file could cause Exim to calculate negative values
543 for the mailbox size or file count. Odd effects could occur as a result.
544 The maildirsize information is now recalculated if the size or filecount
547 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
548 support for a long time. Removed HAVE_SYS_VFS_H.
550 18. Updated exipick to current release from John Jetmore.
552 19. Allow an empty sender to be matched against a lookup in an address list.
553 Previously the only cases considered were a regular expression, or an
556 20. Exim went into a mad DNS lookup loop when doing a callout where the
557 host was specified on the transport, if the DNS lookup yielded more than
560 21. The RFC2047 encoding function was originally intended for short strings
561 such as real names; it was not keeping to the 75-character limit for
562 encoded words that the RFC imposes. It now respects the limit, and
563 generates multiple encoded words if necessary. To be on the safe side, I
564 have increased the buffer size for the ${rfc2047: expansion operator from
567 22. Failure to deliver a bounce message always caused it to be frozen, even if
568 there was an errors_to setting on the router. The errors_to setting is now
571 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
572 canonical form (fully expanded) before being placed in
573 $sender_host_address.
575 24. Updated eximstats to version 1.33
577 25. Include certificate and key file names in error message when GnuTLS fails
578 to set them up, because the GnuTLS error message doesn't include the name
579 of the failing file when there is a problem reading it.
581 26. Expand error message when OpenSSL has problems setting up cert/key files.
584 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
585 (this can affect the format of dates).
587 28. exim_tidydb, when checking for the continued existence of a message for
588 which it has found a message-specific retry record, was not finding
589 messages that were in split spool directories. Consequently, it was
590 deleting retry records that should have stayed in existence.
592 29. eximstats updated to version 1.35
593 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
594 1.35 - bugfix such that pie charts by volume are generated correctly
596 30. The SPA authentication driver was not abandoning authentication and moving
597 on to the next authenticator when an expansion was forced to fail,
598 contradicting the general specification for all authenticators. Instead it
599 was generating a temporary error. It now behaves as specified.
601 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
602 (the order specifies the preference for clients). The order is now AES256,
603 AES128, 3DES, ARCFOUR128.
605 31. Small patch to Sieve code - explicitly set From: when generating an
608 32. Exim crashed if a remote delivery caused a very long error message to be
609 recorded - for instance if somebody sent an entire SpamAssassin report back
610 as a large number of 550 error lines. This bug was coincidentally fixed by
611 increasing the size of one of Exim's internal buffers (big_buffer) that
612 happened as part of the Exiscan merge. However, to be on the safe side, I
613 have made the code more robust (and fixed the comments that describe what
616 33. Some experimental protocols are using DNS PTR records for new purposes. The
617 keys for these records are domain names, not reversed IP addresses. The
618 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
619 leaves it alone. Component reversal etc. now happens only for IP addresses.
622 34. The host_aton() function is supposed to be passed a string that is known
623 to be a valid IP address. However, in the case of IPv6 addresses, it was
624 not checking this. This is a hostage to fortune. Exim now panics and dies
625 if the condition is not met. A case was found where this could be provoked
626 from a dnsdb PTR lookup with an IPv6 address that had more than 8
627 components; fortuitously, this particular loophole had already been fixed
628 by change 4.50/55 or 4.44/33 above.
630 If there are any other similar loopholes, the new check in host_aton()
631 itself should stop them being exploited. The report I received stated that
632 data on the command line could provoke the exploit when Exim was running as
633 exim, but did not say which command line option was involved. All I could
634 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
638 35. There was a buffer overflow vulnerability in the SPA authentication code
639 (which came originally from the Samba project). I have added a test to the
640 spa_base64_to_bits() function which I hope fixes it.
643 36. The daemon start-up calls getloadavg() while still root for those OS that
644 need the first call to be done as root, but it missed one case: when
645 deliver_queue_load_max is set with deliver_drop_privilege. This is
646 necessary for the benefit of the queue runner, because there is no re-exec
647 when deliver_drop_privilege is set.
649 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
650 was in use, was not putting the data itself into the right store pool;
651 consequently, it could be overwritten for a subsequent message in the same
652 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
655 38. Sometimes the final signoff response after QUIT could fail to get
656 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
657 < 0 before doing a fflush(). This bug looks as though it goes back to the
658 introduction of TLS in release 3.20, but "sometimes" must have been rare
659 because the tests only now provoked it.
665 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
666 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
667 Now it returns an integer. A number of calls were still expecting a T/F
668 return. Fortuitously, in all cases, the tests worked in OK situations,
669 which is the norm. However, things would have gone wrong on any write
670 failures on the smtp file descriptor. This function is used when sending
671 messages over SMTP and also when doing verify callouts.
673 2. When Exim is called to do synchronous delivery of a locally submitted
674 message (the -odf or -odi options), it no longer closes stderr before doing
677 3. Implemented the mua_wrapper option.
679 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
681 5. Implemented the functions header_remove(), header_testname(),
682 header_add_at_position(), and receive_remove_recipient(), and exported them
685 6. If an ACL "warn" statement specified the addition of headers, Exim already
686 inserted X-ACL-Warn: at the start if there was no header name. However, it
687 was not making this test for the second and subsequent header lines if
688 there were newlines in the string. This meant that an invalid header could
689 be inserted if Exim was badly configured.
691 7. Allow an ACL "warn" statement to add header lines at the start or after all
692 the Received: headers, as well as at the end.
694 8. Added the rcpt_4xx retry error code.
696 9. Added postmaster_mailfrom=xxx to callout verification option.
698 10. Added mailfrom=xxxx to the callout verification option, for verify=
701 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
702 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
704 12. Inserted some casts to stop certain compilers warning when using pointer
705 differences as field lengths or precisions in printf-type calls (mostly
706 affecting debugging statements).
708 13. Added optional readline() support for -be (dynamically loaded).
710 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
711 same clock tick as a message's arrival, so that its received time was the
712 same as the "first fail" time on the retry record, and that message
713 remained on the queue past the ultimate address timeout, every queue runner
714 would try a delivery (because it was past the ultimate address timeout) but
715 after another failure, the ultimate address timeout, which should have then
716 bounced the address, did not kick in. This was a "< instead of <=" error;
717 in most cases the first failure would have been in the next clock tick
718 after the received time, and all would be well.
720 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
721 being recognized when the domain list was tested by the match_domain
722 condition in an expansion string.
724 16. Added the ${str2b64: operator.
726 17. Exim was always calling setrlimit() to set a large limit for the number of
727 processes, without checking whether the existing limit was already
728 adequate. (It did check for the limit on file descriptors.) Furthermore,
729 errors from getrlimit() and setrlimit() were being ignored. Now they are
730 logged to the main and panic logs, but Exim does carry on, to try to do its
731 job under whatever limits there are.
733 18. Imported PCRE 5.0.
735 19. Trivial typo in log message " temporarily refused connection" (the leading
738 20. If the log selector return_path_on_delivery was set and an address was
739 redirected to /dev/null, the delivery process crashed because it assumed
740 that a return path would always be set for a "successful" delivery. In this
741 case, the whole delivery is bypassed as an optimization, and therefore no
744 21. Internal re-arrangement: the function for sending a challenge and reading
745 a response while authentication was assuming a zero-terminated challenge
746 string. It's now changed to take a pointer and a length, to allow for
747 binary data in such strings.
749 22. Added the cyrus_sasl authenticator (code supplied by MBM).
751 23. Exim was not respecting finduser_retries when seeking the login of the
752 uid under which it was called; it was always trying 10 times. (The default
753 setting of finduser_retries is zero.) Also, it was sleeping after the final
754 failure, which is pointless.
756 24. Implemented tls_on_connect_ports.
758 25. Implemented acl_smtp_predata.
760 26. If the domain in control=submission is set empty, Exim assumes that the
761 authenticated id is a complete email address when it generates From: or
762 Sender: header lines.
764 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
765 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
766 chown and chgrp in /bin and hostname in /usr/bin.
768 28. Exim was keeping the "process log" file open after each use, just as it
769 does for the main log. This opens the possibility of it remaining open for
770 long periods when the USR1 signal hits a daemon. Occasional processlog
771 errors were reported, that could have been caused by this. Anyway, it seems
772 much more sensible not to leave this file open at all, so that is what now
775 29. The long-running daemon process does not normally write to the log once it
776 has entered its main loop, and it closes the log before doing so. This is
777 so that log files can straightforwardly be renamed and moved. However,
778 there are a couple of unusual error situations where the daemon does write
779 log entries, and I had neglected to close the log afterwards.
781 30. The text of an SMTP error response that was received during a remote
782 delivery was being truncated at 512 bytes. This is too short for some of
783 the long messages that one sometimes sees. I've increased the limit to
786 31. It is now possible to make retry rules that apply only when a message has a
787 specific sender, in particular, an empty sender.
789 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
790 it possible to be selective about when SMTP synchronization is enforced.
792 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
794 32. Implemented hosts_connection_nolog.
796 33. Added an ACL for QUIT.
798 34. Setting "delay_warning=" to disable warnings was not working; it gave a
801 35. Added mailbox_size and mailbox_filecount to appendfile.
803 36. Added control = no_multiline_responses to ACLs.
805 37. There was a bug in the logic of the code that waits for the clock to tick
806 in the case where the clock went backwards by a substantial amount such
807 that the microsecond fraction of "now" was more than the microsecond
808 fraction of "then" (but the whole seconds number was less).
810 38. Added support for the libradius Radius client library this is found on
811 FreeBSD (previously only the radiusclient library was supported).
817 1. When certain lookups returned multiple values in the form name=value, the
818 quoting of the values was not always being done properly. Specifically:
819 (a) If the value started with a double quote, but contained no whitespace,
821 (b) If the value contained whitespace other than a space character (i.e.
822 tabs or newlines or carriage returns) it was not quoted.
823 This fix has been applied to the mysql and pgsql lookups by writing a
824 separate quoting function and calling it from the lookup code. The fix
825 should probably also be applied to nisplus, ibase and oracle lookups, but
826 since I cannot test any of those, I have not disturbed their existing code.
828 2. A hit in the callout cache for a specific address caused a log line with no
829 reason for rejecting RCPT. Now it says "Previous (cached) callout
830 verification failure".
832 3. There was an off-by-one bug in the queryprogram router. An over-long
833 return line was truncated at 256 instead of 255 characters, thereby
834 overflowing its buffer with the terminating zero. As well as fixing this, I
835 have increased the buffer size to 1024 (and made a note to document this).
837 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
838 when Exim is waiting for an SMTP response from a remote server, Exim
839 restarts its select() call on the socket, thereby resetting its timeout.
840 This is not a problem when such interrupts are rare. Somebody set up a cron
841 job to run exiwhat every 2 minutes, which is less than the normal select()
842 timeout (5 or 10 minutes). This meant that the select() timeout never
843 kicked in because it was always reset. I have fixed this by comparing the
844 time when an interrupt arrives with the time at the start of the first call
845 to select(). If more time than the timeout has elapsed, the interrupt is
846 treated as a timeout.
848 5. Some internal re-factoring in preparation for the addition of Sieve
849 extensions (by MH). In particular, the "personal" test is moved to a
850 separate function, and given an option for scanning Cc: and Bcc: (which is
851 not set for Exim filters).
853 6. When Exim created an email address using the login of the caller as the
854 local part (e.g. when creating a From: or Sender: header line), it was not
855 quoting the local part when it contained special characters such as @.
857 7. Installed new OpenBSD configuration files.
859 8. Reworded some messages for syntax errors in "and" and "or" conditions to
860 try to make them clearer.
862 9. Callout options, other than the timeout value, were being ignored when
863 verifying sender addresses in header lines. For example, when using
865 verify = header_sender/callout=no_cache
867 the cache was (incorrectly) being used.
869 10. Added a missing instance of ${EXE} to the exim_install script; this affects
870 only the Cygwin environment.
872 11. When return_path_on_delivery was set as a log selector, if different remote
873 addresses in the same message used different return paths and parallel
874 remote delivery occurred, the wrong values would sometimes be logged.
875 (Whenever a remote delivery process finished, the return path value from
876 the most recently started remote delivery process was logged.)
878 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
879 lines when AUTH and/or TLS are in use. This is the "received protocol"
880 field. Exim used to use "asmtp" for authenticated SMTP, without any
881 indication (in the protocol name) for TLS use. Now it follows the RFC and
882 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
883 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
884 names appear in log lines as well as in Received: header lines.
886 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
887 extensions, and comparison tests, and to fix some bugs.
889 14. Changes to the "personal" filter test:
891 (1) The test was buggy in that it was just doing the equivalent of
892 "contains" tests on header lines. For example, if a user's address was
893 anne@some.where, the "personal" test would incorrectly be true for
895 To: susanne@some.where
897 This test is now done by extracting each address from the header in turn,
898 and checking the entire address. Other tests that are part of "personal"
899 are now done using regular expressions (for example, to check local parts
900 of addresses in From: header lines).
902 (2) The list of non-personal local parts in From: addresses has been
903 extended to include "listserv", "majordomo", "*-request", and "owner-*",
904 taken from the Sieve specification recommendations.
906 (3) If the message contains any header line starting with "List-" it is
907 treated as non-personal.
909 (4) The test for "circular" in the Subject: header line has been removed
910 because it now seems ill-conceived.
912 15. Minor typos in src/EDITME comments corrected.
914 16. Installed latest exipick from John Jetmore.
916 17. If headers_add on a router specified a text string that was too long for
917 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
918 of string_sprintf() is now avoided.
920 18. $message_body_size was not set (it was always zero) when running the DATA
921 ACL and the local_scan() function.
923 19. For the "mail" command in an Exim filter, no default was being set for
924 the once_repeat time, causing a random time value to be used if "once" was
925 specified. (If the value happened to be <= 0, no repeat happened.) The
926 default is now 0s, meaning "never repeat". The "vacation" command was OK
927 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
928 (I found it when inspecting the code).
930 20. There is now an overall timeout for performing a callout verification. It
931 defaults to 4 times the callout timeout, which applies to individual SMTP
932 commands during the callout. The overall timeout applies when there is more
933 than one host that can be tried. The timeout is checked before trying the
934 next host. This prevents very long delays if there are a large number of
935 hosts and all are timing out (e.g. when the network connections are timing
936 out). The value of the overall timeout can be changed by specifying an
937 additional sub-option for "callout", called "maxwait". For example:
939 verify = sender/callout=5s,maxwait=20s
941 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
942 to the end before writing, but this should make it even safer).
944 22. Exim was forgetting that it had advertised PIPELINING for the second and
945 subsequent messages on an SMTP connection. It was also not resetting its
946 memory on STARTTLS and an internal HELO.
948 23. When Exim logs an SMTP synchronization error within a session, it now
949 records whether PIPELINING has been advertised or not.
951 24. Added 3 instances of "(long int)" casts to time_t variables that were being
952 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
953 rather than long int.
955 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
957 26. Added the never_mail option to autoreply.
963 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
964 crash if the getsockname() call failed; this can happen if a connection is
965 closed very soon after it is established. The problem was simply in the
966 order in which certain operations were done, causing Exim to try to write
967 to the SMTP stream before it had set up the file descriptor. The bug has
968 been fixed by making things happen in the correct order.
974 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
975 before the connection was closed, thus losing the rejection response.
977 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
978 some early Solaris releases, but causes trouble in current releases where
979 socklen_t is defined.
981 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
984 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
987 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
988 that releasing the top of it at the end releases what was used for sub-
989 expansions (unless the block got too big). However, discard this block if
990 the first thing is a variable or header, so that we can use its block when
991 it is dynamic (useful for very large $message_headers, for example).
993 6. Lookups now cache *every* query, not just the most recent. A new, separate
994 store pool is used for this. It can be recovered when all lookup caches are
995 flushed. Lookups now release memory at the end of their result strings.
996 This has involved some general refactoring of the lookup sources.
998 7. Some code has been added to the store_xxx() functions to reduce the amount
999 of flapping under certain conditions.
1001 8. log_incoming_interface used to affect only the <= reception log lines. Now
1002 it causes the local interface and port to be added to several more SMTP log
1003 lines, for example "SMTP connection from", and rejection lines.
1005 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1007 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1009 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1010 could be overwritten at the end of the current message (or the start of a
1011 new message if it was set in a HELO ACL). The value is now preserved for
1012 the duration of the SMTP connection.
1014 12. If a transport had a headers_rewrite setting, and a matching header line
1015 contained an unqualified address, that address was qualified, even if it
1016 did not match any rewriting rules. The underlying bug was that the values
1017 of the flags that permit the existence of unqualified sender and recipient
1018 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1019 non-local messages, and by -bnq for local messages) were not being
1020 preserved with the message after it was received.
1022 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1023 "next input=" as part of the text comprising the host identity instead of
1024 the correct text. The code was using the same buffer for two different
1025 strings. However, depending on which order the printing function evaluated
1026 its arguments, the bug did not always show up. Under Linux, for example, my
1027 test suite worked just fine.
1029 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1030 doesn't work with some older versions of Perl. It has been changed to "my",
1031 which in any case is probably the better facility to use.
1033 15. A really picky compiler found some instances of statements for creating
1034 error messages that either had too many or two few arguments for the format
1037 16. The size of the buffer for calls to the DNS resolver has been increased
1038 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1039 for addresses that have a lot of PTR records. This alleviates a problem; it
1040 does not fully solve it.
1042 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1043 buffer now truncates the list and logs the incident, which is the same
1044 action as happens when Exim is looking up a host name and its aliases.
1045 Previously in this situation something unpredictable would happen;
1046 sometimes it was "internal error: store_reset failed".
1048 18. If a server dropped the connection unexpectedly when an Exim client was
1049 using GnuTLS and trying to read a response, the client delivery process
1050 crashed while trying to generate an error log message.
1052 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1053 string, for example:
1055 warn message = H1: something\nH2: something
1057 the text was added as a single header line from Exim's point of view
1058 though it ended up OK in the delivered message. However, searching for the
1059 second and subsequent header lines using $h_h2: did not work. This has been
1060 fixed. Similarly, if a system filter added multiple headers in this way,
1061 the routers could not see them.
1063 20. Expanded the error message when iplsearch is called with an invalid key to
1064 suggest using net-iplsearch in a host list.
1066 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1067 ACLs are no longer actually imposed (and a message to that effect is
1070 22. If a "gecos" field in a passwd entry contained escaped characters, in
1071 particular, if it contained a \" sequence, Exim got it wrong when building
1072 a From: or a Sender: header from that name. A second bug also caused
1073 incorrect handling when an unquoted " was present following a character
1074 that needed quoting.
1076 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1077 was not being matched caselessly.
1079 24. Arranged for all hyphens in the exim.8 source to be escaped with
1082 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1083 properly. Recipient callout cache records were still being keyed to include
1084 the sender, even when use_sender was set false. This led to far more
1085 callouts that were necessary. The sender is no longer included in the key
1086 when use_sender is false.
1088 26. Added "control = submission" modifier to ACLs.
1090 27. Added the ${base62d: operator to decode base 62 numbers.
1092 28. dnsdb lookups can now access SRV records.
1094 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1095 the configuration file.
1097 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1098 (-v) mode. This makes the output for a verbose queue run more intelligible.
1100 31. Added a use_postmaster feature to recipient callouts.
1102 32. Added the $body_zerocount variable, containing the number of binary zero
1103 bytes in the message body.
1105 33. The time of last modification of the "new" subdirectory is now used as the
1106 "mailbox time last read" when there is a quota error for a maildir
1109 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1111 35. Added +ignore_unknown as a special item in host lists.
1113 36. Code for decoding IPv6 addresses in host lists is now included, even if
1114 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1115 address was recognized as an IP address, but was then not correctly decoded
1116 into binary, causing unexpected and incorrect effects when compared with
1123 1. Very minor rewording of debugging text in manualroute to say "list of
1124 hosts" instead of "hostlist".
1126 2. If verify=header_syntax was set, and a header line with an unqualified
1127 address (no domain) and a large number of spaces between the end of the
1128 name and the colon was received, the reception process suffered a buffer
1129 overflow, and (when I tested it) crashed. This was caused by some obsolete
1130 code that should have been removed. The fix is to remove it!
1132 3. When running in the test harness, delay a bit after writing a bounce
1133 message to get a bit more predictability in the log output.
1135 4. Added a call to search_tidyup() just before forking a reception process. In
1136 theory, someone could use a lookup in the expansion of smtp_accept_max_
1137 per_host which, without the tidyup, could leave open a database connection.
1139 5. Added the variables $recipient_data and $sender_data which get set from a
1140 lookup success in an ACL "recipients" or "senders" condition, or a router
1141 "senders" option, similar to $domain_data and $local_part_data.
1143 6. Moved the writing of debug_print from before to after the "senders" test
1146 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1147 problems for message scanning, either using a data ACL, or using
1148 local_scan() because the Received: header was not generated till after they
1149 were called (in order to set the time as the time of reception completion).
1150 I have revised the way this works. The header is now generated after the
1151 body is received, but before the ACL or local_scan() are called. After they
1152 are run, the timestamp in the header is updated.
1158 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1159 before starting a queue runner without re-exec. This happened only when
1160 deliver_drop_privilege was set or when the Exim user was set to root. The
1161 effect of the bug was that timeouts during subsequent deliveries caused
1162 crashes instead of being properly handled. The handler is now left at its
1163 default (and expected) setting.
1165 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1166 message, again when deliver_drop_privilege is set or Exim is run as root.
1167 The bug described in (1) was not present in this case, but the tidying up
1168 of the other signals was missing. I have made the two cases consistent.
1170 3. The ignore_target_hosts setting on a manualroute router was being ignored
1171 for hosts that were looked up using the /MX notation.
1173 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1176 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1177 operated on the sender address. After changing the $sender_address to <>
1178 for the sender address verify, Exim was re-instated it as the original
1179 (before rewriting) address, but remembering that it had rewritten it, so it
1180 wasn't rewriting it again. This bug also had the effect of breaking the
1181 sender address verification caching when the sender address was rewritten.
1183 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1184 This has been changed so that if the ip literal address matches
1185 ignore_target_hosts, the router declines.
1187 7. Added expansion conditions match_domain, match_address, and match_local_
1188 part (NOT match_host).
1190 8. The placeholder for the Received: header didn't have a length field set.
1192 9. Added code to Exim itself and to exim_lock to test for a specific race
1193 condition that could lead to file corruption when using MBX delivery. The
1194 issue is with the lockfile that is created in /tmp. If this file is removed
1195 after a process has opened it but before that process has acquired a lock,
1196 there is the potential for a second process to recreate the file and also
1197 acquire a lock. This could lead to two Exim processes writing to the file
1198 at the same time. The added code performs the same test as UW imapd; it
1199 checks after acquiring the lock that its file descriptor still refers to
1200 the same named file.
1202 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1203 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1206 11. Added the smtp_active_hostname option. If used, this will typically be made
1207 to depend on the incoming interface address. Because $interface_address is
1208 not set up until the daemon has forked a reception process, error responses
1209 that can happen earlier (such as "too many connections") no longer contain
1212 12. If an expansion in a condition on a "warn" statement fails because a lookup
1213 defers, the "warn" statement is abandoned, and the next ACL statement is
1214 processed. Previously this caused the whole ACL to be aborted.
1216 13. Added the iplsearch lookup type.
1218 14. Added ident_timeout as a log selector.
1220 15. Added tls_certificate_verified as a log selector.
1222 16. Added a global option tls_require_ciphers (compare the smtp transport
1223 option of the same name). This controls incoming TLS connections.
1225 17. I finally figured out how to make tls_require_ciphers do a similar thing
1226 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1227 before starting the TLS session.
1229 18. Tabs are now shown as \t in -bP output.
1231 19. If the log selector return_path_on_delivery was set, Exim crashed when
1232 bouncing a message because it had too many Received: header lines.
1234 20. If two routers both had headers_remove settings, and the first one included
1235 a superfluous trailing colon, the final name in the first list and the
1236 first name in the second list were incorrectly joined into one item (with a
1237 colon in the middle).
1243 1. Added -C and -D options to the exinext utility, mainly to make it easier
1244 to include in the automated testing, but these could be helpful when
1245 multiple configurations are in use.
1247 2. The exinext utility was not formatting the output nicely when there was
1248 an alternate port involved in the retry record key, nor when there was a
1249 message id as well (for retries that were specific to a specific message
1250 and a specific host). It was also confused by IPv6 addresses, because of
1251 the additional colons they contain. I have fixed the IPv4 problem, and
1252 patched it up to do a reasonable job for IPv6.
1254 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1255 delivery, the log line now contains "pipelined" if PIPELINING was used.
1257 4. An SMTP transport process used to panic and die if the bind() call to set
1258 an explicit outgoing interface failed. This has been changed; it is now
1259 treated in the same way as a connect() failure.
1261 5. A reference to $sender_host_name in the part of a conditional expansion
1262 that was being skipped was still causing a DNS lookup. This no longer
1265 6. The def: expansion condition was not recognizing references to header lines
1266 that used bh_ and bheader_.
1268 7. Added the _cache feature to named lists.
1270 8. The code for checking quota_filecount in the appendfile transport was
1271 allowing one more file than it should have been.
1273 9. For compatibility with Sendmail, the command line option
1281 and sets the incoming protocol and host name (for trusted callers). The
1282 host name and its colon can be omitted when only the protocol is to be set.
1283 Note the Exim already has two private options, -pd and -ps, that refer to
1284 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1285 "s", but I don't think that's a major issue.
1287 10. A number of refactoring changes to the code, none of which should affect
1290 (a) The number of logging options was getting close to filling up the
1291 32-bit word that was used as a bit map. I have split them into two classes:
1292 those that are passed in the argument to log_write(), and those that are
1293 only ever tested independently outside of that function. These are now in
1294 separate 32-bit words, so there is plenty of room for expansion again.
1295 There is no change in the user interface or the logging behaviour.
1297 (b) When building, for example, log lines, the code previously used a
1298 macro that called string_cat() twice, in order to add two strings. This is
1299 not really sufficiently general. Furthermore, there was one instance where
1300 it was actually wrong because one of the argument was used twice, and in
1301 one call a function was used. (As it happened, calling the function twice
1302 did not affect the overall behaviour.) The macro has been replaced by a
1303 function that can join an arbitrary number of extra strings onto a growing
1306 (c) The code for expansion conditions now uses a table and a binary chop
1307 instead of a serial search (which was left over from when there were very
1308 few conditions). Also, it now recognizes conditions like "pam" even when
1309 the relevant support is not compiled in: a suitably worded error message is
1310 given if an attempt is made to use such a condition.
1312 11. Added ${time_interval:xxxxx}.
1314 12. A bug was causing one of the ddress fields not to be passed back correctly
1315 from remote delivery subprocesses. The field in question was not being
1316 subsequently used, so this caused to problems in practice.
1318 13. Added new log selectors queue_time and deliver_time.
1320 14. Might have fixed a bug in maildirsizefile handling that threw up
1321 "unexpected character" debug warnings, and recalculated the data
1322 unnecessarily. In any case, I expanded the warning message to give more
1325 15. Added the message "Restricted characters in address" to the statements in
1326 the default ACL that block characters like @ and % in local parts.
1328 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1329 Three changes have been made:
1331 (a) There was a serious bug; a negative response to MAIL caused the whole
1332 recipient domain to be cached as invalid, thereby blocking all messages
1333 to all local parts at the same domain, from all senders. This bug has
1334 been fixed. The domain is no longer cached after a negative response to
1335 MAIL if the sender used is not empty.
1337 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1340 (c) A new callout option, "use_sender" has been added for people who want
1341 the modified behaviour.
1347 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1350 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1351 indeed breaks things for older releases.
1353 3. Added additional logging to the case where there is a problem reading data
1354 from a filter that is running in a subprocess using a pipe, in order to
1355 try to track down a specific problem.
1357 4. Testing facility fudge: when running in the test harness and attempting
1358 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1359 getting "No route to host". Convert this to a timeout.
1361 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1364 6. Some OS don't have socklen_t but use size_t instead. This affects the
1365 fifth argument of getsockopt() amongst other things. This is now
1366 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1367 can be set for individual OS. I have set it for SunOS5, OSF1, and
1368 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1369 some earlier ones do not.
1371 7. Change 4.30/15 was not doing the test caselessly.
1373 8. The standard form for an IPv6 address literal was being rejected by address
1374 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1375 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1376 this, as well as the form without the "IPv6" on the front (but only when
1377 address literals are enabled, of course).
1379 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1381 10. Exim crashed if a message with an empty sender address specified by -f
1382 encountered a router with an errors_to setting. This could be provoked only
1383 by a command such as
1387 where an empty string was supplied; "<>" did not hit this bug.
1389 11. Installed PCRE release 4.5.
1391 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1392 remained set. It is now erased.
1394 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1395 times from message ids (which are base 36 rather than the normal 62).
1397 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1398 were being counted as actual protocol errors, and logged if the log
1399 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1400 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1401 and DATA following a set of rejected RCPTs do not count as protocol errors.
1402 In other words, Exim assumes they were pipelined, though this may not
1403 actually be the case. Of course, in all cases the client gets an
1404 appropriate error code.
1406 15. If a lookup fails in an ACL condition, a message about the failure may
1407 be available; it is used if testing the ACL cannot continue, because most
1408 such messages specify what the cause of the deferral is. However, some
1409 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1410 that caused an old message to be retained and used if a later statement
1411 caused a defer, replacing the real cause of the deferral.
1413 16. If an IP address had so many PTR records that the DNS lookup buffer
1414 was not large enough to hold them, Exim could crash while trying to process
1415 the truncated data. It now detects and logs this case.
1417 17. Further to 4.21/58, another change has been made: if (and only if) the
1418 first line of a message (the first header line) ends with CRLF, a bare LF
1419 in a subsequent header line has a space inserted after it, so as not to
1420 terminate the header.
1422 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1423 unnecessarily, used atoi() instead of strtol(), and didn't check the
1424 termination when getting file sizes from file names by regex.
1426 19. Completely re-implemented the support for maildirsize files, in the light
1427 of a number of problems with the previous contributed implementation
1428 (4.30/29). In particular:
1430 . If the quota is zero, the maildirsize file is maintained, but no quota is
1433 . If the maildir directory does not exist, it is created before any attempt
1434 to write a maildirsize file.
1436 . The quota value in the file is just a cache; if the quota is changed in
1437 the transport, the new value overrides.
1439 . A regular expression is available for excluding directories from the
1442 20. The autoreply transport checks the characters in options that define the
1443 message's headers; it allows continued headers, but it was checking with
1444 isspace() after an embedded newline instead of explicitly looking for a
1447 21. If all the "regular" hosts to which an address was routed had passed their
1448 expiry times, and had not reached their retry times, the address was
1449 bounced, even if fallback hosts were defined. Now Exim should go on to try
1452 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1453 equivalent code in the SMTP transport. Some hosts send humungous responses
1454 to HELO/EHLO, more than 1024 it seems.
1456 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1457 gives compiler warnings in some environments. I've now done it "properly",
1460 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1461 (because of problems with the built-in one) was declared to return uschar *
1462 instead of char *, causing compiler failure.
1464 25. Fixed a file descriptor leak when processing alias/forward files.
1466 26. Fixed a minor format string issue in dbfn.c.
1468 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1470 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1471 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1474 29. When a sender address is verified, it is cached, to save repeating the test
1475 when there is more than one recipient in a message. However, when the
1476 verification involves a callout, it is possible for different callout
1477 options to be set for different recipients. It is too complicated to keep
1478 track of this in the cache, so now Exim always runs a verification when a
1479 callout is required, relying on the callout cache for the optimization.
1480 The overhead is duplication of the address routing, but this should not be
1483 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1484 to be verified with callout=postmaster, and the main callout worked but the
1485 postmaster check failed, the verification correctly failed. However, if a
1486 subsequent RCPT command asked for sender verification *without* the
1487 postmaster check, incorrect caching caused this verification also to fail,
1490 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1491 it was not caching the DNS options (qualify_single, search_parents) that
1492 were used when the lookup failed. A subsequent lookup with different
1493 options therefore always gave the same answer, though there were cases
1494 where it should not have. (Example: a "domains = !$mx_any" option on a
1495 dnslookup router: the "domains" option is always processed without any
1496 widening, but the router might have qualify_single set.) Now Exim uses the
1497 cached value only when the same options are set.
1499 32. Added John Jetmore's "exipick" utility to the distribution.
1501 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1502 than a timeout (e.g. a certificate is required, and is not provided), an
1503 Exim server now closes the connection immediately. Previously it waited for
1504 the client to close - but if the client is SSL, it seems that they each
1505 wait for each other, leading to a delay before one of them times out.
1507 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1508 maintained 0.8.x compatibility because I don't think many are using it, and
1509 it is clearly obsolete.
1511 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1514 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1515 client certificate was expired. A simple patch fixes this, though I don't
1516 understand the full logic of why the verify callback is called multiple
1519 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1520 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1521 which causes problems with some clients (such as the Certicom SSL Plus
1522 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1523 disables the coutermeasure allowing Eudora to connect."
1525 38. Exim was not checking that a write() to a log file succeeded. This could
1526 lead to Bad Things if a log got too big, in particular if it hit a file
1527 size limit. Exim now panics and dies if it cannot write to a log file, just
1528 as it does if it cannot open a log file.
1530 39. Modified OS/Makefile-Linux so that it now contains
1532 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1534 The two -D definitions ensure that Exim is compiled with large file
1535 support, which makes it possible to handle log files that are bigger than
1538 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1539 instance) a domain was checked against a named list that involved a lookup,
1540 causing $domain_data to be set, then another domain was checked against the
1541 same list, then the first domain was re-checked, the value of $domain_data
1542 after the final check could be wrong. In particular, if the second check
1543 failed, it could be set empty. This bug probably also applied to
1546 41. The strip_trailing_dot option was not being applied to the address given
1547 with the -f command-line option.
1549 42. The code for reading a message's header from the spool was incrementing
1550 $received_count, but never initializing it. This meant that the value was
1551 incorrect (doubled) while delivering a message in the same process in which
1552 it was received. In the most common configuration of Exim, this never
1553 happens - a fresh exec is done - but it can happen when
1554 deliver_drop_privilege is set.
1556 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1557 it now includes up to 150 characters of the unexpected data in the log
1560 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1561 and building data strings. The size of both of these buffers was 10 000
1562 bytes - far larger than anybody would *ever* want, thought I. Needless to
1563 say, somebody hit the limit. I have increased the maximum line length to
1564 20 000 and the maximum data length of concatenated lines to 100 000. I have
1565 also fixed two bugs, because there was no checking on these buffers. Tsk,
1566 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1567 buffer is too small.
1569 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1570 lsearch lookups. Now it does.
1572 46. When parsing a route_list item in a manualroute router, a fixed-length
1573 buffer was used for the list of hosts. I made this 1024 bytes long,
1574 thinking that nobody would ever have a list of hosts that long. Wrong.
1575 Somebody had a whole pile of complicated expansion conditions, and the
1576 string was silently truncated, leading to an expansion error. It turns out
1577 that it is easier to change to an unlimited length (owing to other changes
1578 that have happened since this code was originally written) than to build
1579 structure for giving a limitation error. The length of the item that
1580 expands into the list of hosts is now unlimited.
1582 47. The lsearch lookup could not handle data where the length of text line was
1583 more than 4095 characters. Such lines were truncated, leading to shortened
1584 data being returned. It should now handle lines of any length.
1586 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1587 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1590 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1591 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1592 no longer happens in comment lines. A list of replacements is now placed
1593 at the head of all of the source files, except those whose only change is
1594 to replace PERL_COMMAND in the very first #! line.
1596 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1597 messages on the queue, with a bottom-up merge sort, using code contributed
1598 by Michael Haardt. This should make operations like -bp somewhat faster on
1599 large queues. It won't affect queue runners, except when queue_run_in_order
1602 51. Installed eximstats 1.31 in the distribution.
1604 52. Added support for SRV lookups to the dnslookup router.
1606 53. If an ACL referred to $message_body or $message_body_end, the value was not
1607 reset for any messages that followed in the same SMTP session.
1609 54. The store-handling optimization for building very long strings was not
1610 differentiating between the different store pools. I don't think this
1611 actually made any difference in practice, but I've tidied it.
1613 55. While running the routers to verify a sender address, $sender_address
1614 was still set to the sender address. This is wrong, because when routing to
1615 send a bounce to the sender, it would be empty. Therefore, I have changed
1616 it so that, while verifying a sender address, $sender_address is set to <>.
1617 (There is no change to what happens when verifying a recipient address.)
1619 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1620 target A or AAAA records (if not already returned) without resetting the
1621 qualify_single or search_parents options of the DNS resolver. These are
1622 inappropriate in this case because the targets of MX and SRV records must
1623 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1624 target that, when qualified, matched something in the local domain. These
1625 two options are now turned off when doing these lookups.
1627 57. It seems that at least some releases of Reiserfs (which does not have the
1628 concept of a fixed number of inodes) returns zero and not -1 for the
1629 number of available inodes. This interacted badly with check_spool_inodes,
1630 which assumed that -1 was the "no such thing" setting. What I have done is
1631 to check that the total number of inodes is greater than zero before doing
1632 the test of how many are available.
1634 58. When a "warn" ACL statement has a log_message modifier, the message is
1635 remembered, and not repeated. This is to avoid a lot of repetition when a
1636 message has many recipients that cause the same warning to be written.
1637 Howewer, Exim was preserving the list of already written lines for an
1638 entire SMTP session, which doesn't seem right. The memory is now reset if a
1639 new message is started.
1641 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1642 debugging output unless log_rewrite was also set.
1644 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1645 of (int)(handle) when we know that handle contains (void *)(-1).
1647 61. The Exim daemon panic-logs an error return when it closes the incoming
1648 connection. However "connection reset by peer" seems to be common, and
1649 isn't really an error worthy of noting specially, so that particular error
1652 62. When Exim is trying to find all the local interfaces, it used to panic and
1653 die if the ioctl to get the interface flags failed. However, it seems that
1654 on at least one OS (Solaris 9) it is possible to have an interface that is
1655 included in the list of interfaces, but for which you get a failure error
1656 for this call. This happens when the interface is not "plumbed" into a
1657 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1658 failure of the "get flags" call assumes that the interface is down.
1660 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1661 makes life easier for people who are doing arithmetic on fields extracted
1662 from dates, where you often get leading zeros that should not be
1663 interpreted as octal.
1665 64. Added qualify_domain to the redirect router, to override the global
1668 65. If a pathologically long header line contained very many addresses (the
1669 report of this problem mentioned 10 000) and each of them was rewritten,
1670 Exim could use up a very large amount of memory. (It kept on making new
1671 copies of the header line as it rewrote, and never released the old ones.)
1672 At the expense of a bit more processing, the header rewriting function has
1673 been changed so that it no longer eats memory in this way.
1675 66. The generation of the Received: header has been moved from the time that a
1676 message starts to be received, to the time that it finishes. The timestamp
1677 in the Received: header should now be very close to that of the <= log
1678 line. There are two side-effects of this change:
1680 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1681 logged header lines no longer include the local Received: line, because
1682 it has not yet been created. The same applies to a copy of the message
1683 that is returned to a non-SMTP sender when a message is rejected.
1685 (b) When a filter file is tested using -bf, no additional Received: header
1686 is added to the test message. After some thought, I decided that this
1689 This change does not affect the value of $received_for. It is still set
1690 after address rewriting, but before local_scan() is called.
1692 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1694 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1695 gave an unhelpful panic error message, and a defer error. I have managed to
1696 change this behaviour so that it now rejects any supplied certificate,
1697 which seems right, as the list of acceptable certificates is empty.
1699 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1700 gave an unhelpful defer error. I have not managed to make this reject any
1701 supplied certificates, but the error message it gives is "no certificate
1702 supplied", which is not helpful.
1704 70. exigrep's output now also includes lines that are not associated with any
1705 message, but which match the given pattern. Implemented by a patch from
1706 Martin Sluka, which also tidied up the Perl a bit.
1708 71. Recipient callout verification, like sender verification, was using <> in
1709 the MAIL FROM command. This isn't really the right thing, since the actual
1710 sender may affect whether the remote host accepts the recipient or not. I
1711 have changed it to use the actual sender in the callout; this means that
1712 the cache record is now keyed on a recipient/sender pair, not just the
1713 recipient address. There doesn't seem to be a real danger of callout loops,
1714 since a callout by the remote host to check the sender would use <>.
1715 [SEE ABOVE: changed after hitting problems.]
1717 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1718 temporary errors. However, in the case of such a code being given after
1719 the end of a data transmission (i.e. after ".") Exim was failing to write
1720 a retry record for the message. (Yes, there was some broken host that was
1721 actually sending 8xx at this point.)
1723 73. An unknown lookup type in a host list could cause Exim to panic-die when
1724 the list was checked. (An example that provoked this was putting <; in the
1725 middle of a list instead of at the start.) If this happened during a DATA
1726 ACL check, a -D file could be left lying around. This kind of configuration
1727 error no longer causes Exim to die; instead it causes a defer errror. The
1728 incident is still logged to the main and panic logs.
1730 74. Buglet left over from Exim 3 conversion. The message "too many messages
1731 in one connection" was written to the rejectlog but not the mainlog, except
1732 when address rewriting (yes!) was being logged.
1734 75. Added write_rejectlog option.
1736 76. When a system filter was run not as root (that is, when system_filter_user
1737 was set), the values of the $n variables were not being returned to the
1738 main process; thus, they were not subsequently available in the $sn
1741 77. Added +return_path_on_delivery log selector.
1743 78. A connection timeout was being treated differently from recipients deferred
1744 when testing hosts_max_try with a message that was older than the host's
1745 retry timeout. (The host should not be counted, thus allowing all hosts to
1746 be tried at least once before bouncing.) This may have been the cause of an
1747 occasionally reported bug whereby a message would remain on the queue
1748 longer than the retry timeout, but would be bounced if a delivery was
1749 forced. I say "may" because I never totally pinned down the problem;
1750 setting up timeout/retry tests is difficult. See also the next item.
1752 79. The ultimate address timeout was not being applied to errors that involved
1753 a combination of host plus message (for example, a timeout on a MAIL
1754 command). When an address resolved to a number of possible hosts, and they
1755 were not all tried for each delivery (e.g. because of hosts_max_try), a
1756 message could remain on the queue longer than the retry timeout.
1758 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1761 81. Fixed an obscure SMTP outgoing bug which required at least the following
1762 conditions: (a) there was another message waiting for the same server;
1763 (b) the server returned 5xx to all RCPT commands in the first message so
1764 that the message was not completed; (c) the server dropped the connection
1765 or gave a negative response to the RSET that Exim sends to abort the
1766 transaction. The observed case was a dropped connection after DATA that had
1767 been sent in pipelining mode. That is, the server had advertised PIPELINING
1768 but was not implementing it correctly. The effect of the bug was incorrect
1769 behaviour, such as trying another host, and this could lead to a crash.
1775 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1776 and daemon.c were passed as pointers to ints; they should have been
1777 pointers to socklen_t variables (which are typically unsigned ints).
1779 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1782 3. Fixed a really odd bug that affected only the testing scheme; patching a
1783 certain fixed string in the binary changed the value of another string that
1784 happened to be identical to the end of the original first string.
1786 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1787 name", it returns that address as the IP address. On some operating
1788 systems (e.g. Solaris), it also passes back the IP address string as the
1789 "host name". However, on others (e.g. Linux), it passes back an empty
1790 string. Exim wasn't checking for this, and was changing the host name to an
1791 empty string, assuming it had been canonicized.
1793 5. Although rare, it is permitted to have more than one PTR record for a given
1794 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1795 all the names associated with an address, because they do in Solaris.
1796 However, it seems that they do not in Linux for data that comes from the
1797 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1798 I found this out when I moved to a new Linux workstation and tried to run
1799 the Exim test suite.
1801 To get round this problem I have changed the code so that it now does its
1802 own call to the DNS to look up PTR records when searching for a host name.
1803 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1804 addresses that are only in /etc/hosts are still found.
1806 This behaviour is, however, controlled by an option called host_lookup_
1807 order, which defaults to "bydns:byaddr". If people want to use the other
1808 order, or indeed, just use one or the other means of lookup, they can
1809 specify it in this variable.
1811 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1812 some operating systems, this comes back from gethostbyaddr() as an empty
1813 string, and this is what Exim used to test for. However, it seems that in
1814 other systems, "." is yielded. Exim now tests for this case too.
1816 7. The values of check_spool_space and check_log_space are now held internally
1817 as a number of kilobytes instead of an absolute number of bytes. If a
1818 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1819 kilobyte. This means that much larger values can be stored.
1821 8. Exim monitor: an attempt to get the action menu when not actually pointing
1822 at a message produces an empty menu entitled "No message selected". This
1823 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1824 no entries in it ("Shell widget menu has zero width and/or height"). So I
1825 have added a single, blank menu entry in this case.
1827 9. Added ${quote_local_part.
1829 10. MIME decoding is now applied to the contents of Subject: header lines when
1832 11. Now that a reference to $sender_host_address automatically causes a reverse
1833 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1834 host lookup before query-style lookups in lists that might use this
1835 variable. This has therefore been abolished, and the "net-" prefix is no
1836 longer necessary for query-style lookups.
1838 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1839 have been a typo for LFLAGS, so it has been changed.
1841 13. The install script calls Exim with "-C /dev/null" in order to find the
1842 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1843 to be output. Howeve, since Exim outputs its version number before the
1844 error, it didn't break the script. It just looked ugly. I fixed this by
1845 always allowing "-C /dev/null" if the caller is root.
1847 14. Ignore overlarge ACL variable number when reading spool file - insurance
1848 against a later release with more variables having written the file.
1850 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1851 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1852 this, as well as the form without the "IPv6" on the front.
1854 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1855 OS/Makefile-Darwin file.
1857 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1858 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1860 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1861 variable. However, some OS use ints for inodes. I've added cast to long int
1862 to get rid of the compiler warning.
1864 19. I had forgotten to lock out "/../" in configuration file names when
1865 ALT_CONFIG_PREFIX was set.
1867 20. Routers used for verification do not need to specify transports. However,
1868 if such a router generated a host list, and callout was configured, Exim
1869 crashed, because it could not find a port number from the (non-existent)
1870 transport. It now assumes port 25 in this circumstance.
1872 21. Added the -t option to exigrep.
1874 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1875 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1876 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1877 them is set, LOOKUP_LSEARCH is forced.
1879 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1881 24. Added sender and host information to the "rejected by local_scan()" log
1882 line; previously there was no indication of these.
1884 25. Added .include_if_exists.
1886 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1887 Linux. It turns out that not all file systems support this. Apparently some
1888 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1889 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1890 not supported on the file descriptor, is now ignored when Exim is trying to
1891 sync a directory. This applies only to Linux.
1893 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1895 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1896 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1899 29. Added support for maildirsize files from supplied patch (modified a bit).
1901 30. The use of :fail: followed by an empty string could lead Exim to respond to
1902 sender verification failures with (e.g.):
1904 550 Verification failed for <xxx>
1905 550 Sender verify failed
1907 where the first response line was missing the '-' that indicates it is not
1908 the final line of the response.
1910 31. The loop for finding the name of the user that called Exim had a hardwired
1911 limit of 10; it now uses the value of finduser_retries, which is used for
1912 all other user lookups.
1914 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1917 33. Exim was neglecting to zero errno before one call of strtol() when
1918 expanding a string and expecting an integer value. On some systems this
1919 resulted in spurious "integer overflow" errors. Also, it was casting the
1920 result into an int without checking.
1922 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1923 did not work. The code looks as if it has *never* worked, though it appears
1924 to have been documented since at least releast 1.62. I have made it work.
1926 35. The "timeout_DNS" error in retry rules, also documented since at least
1927 1.62, also never worked. As it isn't clear exactly what this means, and
1928 clearly it isn't a major issue, I have abolished the feature by treating it
1929 as "timeout", and writing a warning to the main and panic logs.
1931 36. The display of retry rules for -brt wasn't always showing the error code
1934 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1935 timeout_connect_A, timeout_connect_MX.
1937 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1938 to the empty sender.
1940 39. The daemon was not analysing the content of -oX till after it had closed
1941 stderr and disconnected from the controlling terminal. This meant that any
1942 syntax errors were only noted on the panic log, and the return code from
1943 the command was 0. By re-arranging the code a little, I've made the
1944 decoding happen first, so such errors now appear on stderr, and the return
1945 code is 1. However, the actual setting up of the sockets still happens in
1946 the disconnected process, so errors there are still only recorded on the
1949 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1950 connections (as happens on some IP stacks) was logged at start up time as
1951 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1952 it from "IPv6 and IPv4", which means that two separate sockets are being
1955 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1956 says whether AF_INET or AF_INET6 was passed as an argument.
1958 42. Exiwhat output was messed up when time zones were included in log
1961 43. Exiwhat now gives more information about the daemon's listening ports,
1962 and whether -tls-on-connect was used.
1964 44. The "port" option of the smtp transport is now expanded.
1966 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1967 silently ignored. Now an error message is written to the main and panic
1970 46. There's a new ACL modifier called "logwrite" which writes to a log file
1971 as soon as it is encountered.
1973 47. Added $local_user_uid and $local_user_gid at routing time.
1975 48. Exim crashed when trying to verify a sender address that was being
1978 49. Exim was recognizing only a space character after ".include". It now also
1979 recognizes a tab character.
1981 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1982 extracting the relevant information from the specification. The man page no
1983 longer contains scrambled data for the -d option, and I've added a section
1984 at the front about calling Exim under different names.
1986 51. Added "extra_headers" argument to the "mail" command in filter files.
1988 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1991 53. Installed eximstats 1.29.
1993 54. Added transport_filter_timeout as a generic transport option.
1995 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1996 Cc: header lines. This was required by RFC 822, but it not required by RFC
1999 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2000 incoming messages that did not have them. Now it does so only if the
2001 message originates locally, that is, if there is no associated remote host
2002 address. When Resent- header lines are present, this applies to the Resent-
2003 lines rather than the non-Resent- lines.
2005 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2006 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2008 58. Messages for configuration errors now include the name of the main
2009 configuration file - useful now that there may be more than one file in a
2010 list (.included file names were always shown).
2012 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2013 for those rare installations that do not start the daemon as root or run it
2014 setuid root. I've cut out the call to initgroups() if the daemon is not
2017 60. The Exim user and group can now be bound into the binary as text strings
2018 that are looked up at the start of Exim's processing.
2020 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2022 62. Added $mailstore_basename variable.
2024 63. Installed patch to sieve.c from Michael Haardt.
2026 64. When Exim failed to open the panic log after failing to open the main log,
2027 the original message it was trying to log was written to stderr and debug
2028 output, but if they were not available (the usual case in production), it
2029 was lost. Now it is written to syslog before the two lines that record the
2030 failures to open the logs.
2032 65. Users' Exim filters run in subprocesses under the user's uid. It is
2033 possible for a "deliver" command or an alias in a "personal" command to
2034 provoke an address rewrite. If logging of address rewriting is configured,
2035 this fails because the process is not running as root or exim. There may be
2036 a better way of dealing with this, but for the moment (because 4.30 needs
2037 to be released), I have disabled address rewrite logging when running a
2038 filter in a non-root, non-exim process.
2044 1. The buildconfig auxiliary program wasn't quoting the value set for
2045 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2046 not defined. This bug was masked in 4.22 by the effect that was fixed in
2049 2. Some messages that were rejected after a message id was allocated were
2050 shown as "incomplete" by exigrep. It no longer does this for messages that
2051 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2053 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2054 have allow_domain_literals set, the ID did not get logged in the <= line.
2055 Domain literals are now always recognized in Message-ID: header lines.
2057 4. The first argument for a ${extract expansion item is the key name or field
2058 number. Leading and trailing spaces in this item were not being ignored,
2059 causing some misleading effects.
2061 5. When deliver_drop_privilege was set, single queue runner processes started
2062 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2063 same command in the process it spins off) were not dropping privilege.
2065 6. When the daemon running as "exim" started a queue runner, it always
2066 re-executed Exim in the spun-off process. This is a waste of effort when
2067 deliver_drop_privilege is set. The new process now just calls the
2068 queue-runner function directly.
2074 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2077 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2078 ignored. Though the use of -oP was forcing the writing of a pid file, it
2079 was always written to the default place.
2081 3. If the message "no IP address found for host xxxx" is generated during
2082 incoming verification, it is now followed by identification of the incoming
2083 connection (so you can more easily find what provoked it).
2085 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2087 5. Added some features to "harden" Exim a bit more against certain attacks:
2089 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2090 be put in Local/Makefile. This is like the never_users runtime option,
2091 but it cannot be overridden. The default setting is "root".
2093 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2094 prefix string with which any file named in a -C command line option
2097 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2098 is retained for -C and -D only if the caller of Exim is root. Without
2099 it, the exim user may also use -C and -D and retain privilege.
2101 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2102 command line option is disabled.
2104 6. Macro names set by the -D option must start with an upper case letter, just
2105 like macro names defined in the configuration file.
2107 7. Added "dereference=" facility to LDAP.
2109 8. Two instances of the typo "uknown" in the source files are fixed.
2111 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2112 the Configure-Makefile script screwed up while processing it.
2114 10. Incorporated PCRE 4.4.
2116 11. The SMTP synchronization check was not operating right at the start of an
2117 SMTP session. For example, it could not catch a HELO sent before the client
2118 waited for the greeting. There is now a check for outstanding input at the
2119 point when the greeting is written. Because of the duplex, asynchronous
2120 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2121 way, but not yet received, when the check is performed.
2123 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2124 on TCP/IP sockets, because this apparently causes some broken clients to
2127 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2128 unchanged) from the Cygwin maintainer.
2130 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2131 was supported instead of testing for mbx. Effectively a typo.
2133 15. The spa authenticator server code was not checking that the input it
2134 received was valid base64.
2136 16. The debug output line for the "set" modifier in ACLs was not showing the
2137 name of the variable that was being set.
2139 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2141 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2142 lookup on its own. Something else was needed to trigger the lookup. For
2143 example, a match in host_lookup or the need for a host name in a host list.
2144 Now, if $sender_host_name is referenced and the host name has not yet been
2145 looked up, a lookup is performed. If the lookup fails, the variable remains
2146 empty, and $host_lookup_failed is set to "1".
2148 19. Added "eqi" as a case-independent comparison operator.
2150 20. The saslauthd authentication condition could segfault if neither service
2151 nor realm was specified.
2153 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2154 error message that was logged was misleading, and incoming SMTP
2155 connections were dropped. The message is now more accurate, and temporary
2156 errors are given to SMTP connections.
2158 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2159 (except RSET) with a 421 error, until QUIT is received. However, it was
2160 failing to send a response to QUIT.
2162 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2163 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2164 ACL is now run after the helo_try_verify_hosts code.
2166 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2167 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2168 case-independent, so other case variants are also recognized). Apparently
2169 some systems use these upper case variants.
2171 25. If more than two messages were waiting for the same host, and a transport
2172 filter was specified for the transport, Exim sent two messages over the
2173 same TCP/IP connection, and then failed with "socket operation on non-
2174 socket" when it tried to send the third.
2176 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2178 27. The extern definition of crypt16() in expand.c was not being excluded when
2179 the OS had its own crypt16() function.
2181 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2182 as a preferred synonym for return_size_limit, both as an option and as an
2185 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2187 30. Changed the default configuration ACL to relax the local part checking rule
2188 for addresses that are not in any local domains. For these addresses,
2189 slashes and pipe symbols are allowed within local parts, but the sequence
2190 /../ is explicitly forbidden.
2192 31. SPA server authentication was not clearing the challenge buffer before
2195 32. log_message in a "warn" ACL statement was writing to the reject log as
2196 well as to the main log, which contradicts the documentation and doesn't
2197 seem right (because no rejection is happening). So I have stopped it.
2199 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2200 However, I am unable to do any testing of this.
2202 34. Fixed an infelicity in the appendfile transport. When checking directories
2203 for a mailbox, to see if any needed to be created, it was accidentally
2204 using path names with one or more superfluous leading slashes; tracing
2205 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2207 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2208 contents are added to the log line that is written for every discarded
2209 recipient. (Previously a log_message setting was ignored.)
2211 36. The ${quote: operator now quotes the string if it is empty.
2213 37. The install script runs exim in order to find its version number. If for
2214 some reason other than non-existence or emptiness, which it checks, it
2215 could not run './exim', it was installing it with an empty version number,
2216 i.e. as "exim-". This error state is now caught, and the installation is
2219 38. An argument was missing from the function that creates an error message
2220 when Exim fails to connect to the socket for saslauthd authentication.
2221 This could cause Exim to crash, or give a corrupted message.
2223 39. Added isip, isip4, and isip6 to ${if conditions.
2225 40. The ACL variables $acl_xx are now saved with the message, and can be
2226 accessed later in routers, transports, and filters.
2228 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2229 strings in the file are not string-expanded.
2231 42. If a MAIL command specified a SIZE value that was too large to fit into an
2232 int variable, the check against message_size_limit failed. Such values are
2233 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2234 day this will have to be increased, but I don't think I want to be around
2235 when emails are that large.
2242 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2243 iconv() is not standard in FreeBSD.
2245 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2246 IPv6 enabled. The observed symptom was a segmentation fault on return from
2247 the function os_common_find_running_interfaces() in src/os.c.
2249 3. In the check_special_case() function in daemon.c I had used "errno" as an
2250 argument name, which causes warnings on some systems. This was basically a
2251 typo, since it was named "eno" in the comments!
2253 4. The code that waits for the clock to tick (at a resolution of some fraction
2254 of a second) so as to ensure message-id uniqueness was always waiting for
2255 at least one whole tick, when it could have waited for less. [This is
2256 almost certainly not relevant at current processor speeds, where it is
2257 unlikely to ever wait at all. But we try to future-proof.]
2259 5. The function that sleeps for a time interval that includes fractions of a
2260 second contained a race. It did not block SIGALRM between setting the
2261 timer, and suspending (a couple of lines later). If the interval was short
2262 and the sigsuspend() was delayed until after it had expired, the suspension
2263 never ended. On busy systems this could lead to processes getting stuck for
2266 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2267 process, before it forks any delivery processes. The open lookup caching
2268 mechanism meant that the open file or database connection was passed into
2269 the delivery process. The problem was that delivery processes always tidy
2270 up cached lookup data. This could cause a problem for the next delivery
2271 process started by the queue runner, because the external queue runner
2272 process does not know about the closure. So the next delivery process
2273 still has data in the lookup cache. In the case of a file lookup, there was
2274 no problem because closing a file descriptor in a subprocess doesn't affect
2275 the parent. However, if the lookup was caching a connection to a database,
2276 the connection was closed, and the second delivery process was likely to
2277 see errors such as "PGSQL: query failed: server closed the connection
2278 unexpectedly". The problem has been fixed by closing all cached lookups
2279 in a queue runner before running a delivery process.
2281 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2282 seem to have the "const" qualifier which it has on other OS. I've
2285 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2286 *of the same type* (client or server) with the same public name that an
2287 error should be diagnosed.
2289 9. When Exim looked up a host name for an IP address, but failed to find the
2290 original IP address when looking up the host name (a safety check), it
2291 output the message "<ip address> does not match any IP for NULL", which was
2292 confusing, to say the least. The bug was that the host name should have
2293 appeared instead of "NULL".
2295 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2296 user that is built into the binary, and the -C or -D options is used, root
2297 privilege is dropped before the configuration file is read. In addition,
2298 logging is switched to stderr instead of the normal log files. If the
2299 configuration then re-defines the Exim user, the unprivileged environment
2300 is probably not what is expected, so Exim logs a panic warning message (but
2303 However, if deliver_drop_privilege is set, the unprivileged state may well
2304 be exactly what is intended, so the warning has been cut out in that case,
2305 and Exim is allowed to try to write to its normal log files.
2311 1. smtp_return_error_details was not giving details for temporary sender
2312 or receiver verification errors.
2314 2. Diagnose a configuration error if two authenticators have the same public
2317 3. Exim used not to create the message log file for a message until the first
2318 delivery attempt. This could be confusing when incoming messages were held
2319 for policy or load reasons. The message log file is now created at the time
2320 the message is received, and an initial "Received" line is written to it.
2322 4. The automatically generated man page for command line options had a minor
2323 bug that caused no ill effects; however, a more serious problem was that
2324 the procedure for building the man page automatically didn't always
2325 operate. Consequently, release 4.20 contains an out-of-date version. This
2326 shouldn't happen again.
2328 5. When building Exim with embedded Perl support, the script that builds the
2329 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2330 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2332 6. The freeze_tell option was not being used for messages that were frozen on
2333 arrival, either by an ACL or by local_scan().
2335 7. Added the smtp_incomplete_transaction log selector.
2337 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2338 was accepting AUTH without a new EHLO.
2340 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2341 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2344 10. Logging of TCP/IP connections (when configured) now happens in the main
2345 daemon process instead of the child process, so that the TCP/IP connection
2346 count is more accurate (but it can never be perfect).
2348 11. The use of "drop" in a nested ACL was not being handled correctly in the
2349 outer ACL. Now, if condition failure induced by the nested "drop" causes
2350 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2351 or "require"), the connection is dropped.
2353 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2354 that yield "discard" can now be used with an "accept" or a "discard" verb,
2355 but an error is generated for any others (because I can't see a useful way
2356 to define what should happen).
2358 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2359 are now processed in the same way as lines in the Exim configuration file.
2360 In particular, continuation lines are supported.
2362 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2364 15. Added -ti meaning -t -i.
2366 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2367 domains, and warn by logging if others are found.
2369 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2370 returned from GIFCONF when Exim is trying to find the list of interfaces on
2371 a host. The code in os.c has been modified to copy each ifreq to an aligned
2372 structure in all cases.
2374 Also, in some cases, the returned ifreq's were being copied to a 'struct
2375 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2376 means the last couple of bytes of an IPv6 address could be chopped if the
2377 ifreq contained only a normal sockaddr (14 bytes storage).
2379 18. Named domain lists were not supported in the hosts_treat_as_local option.
2380 An entry such as +xxxx was not recognized, and was treated as a literal
2383 19. Ensure that header lines added by a DATA ACL are included in the reject log
2384 if the ACL subsequently rejects the message.
2386 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2387 MD5 (which is deprecated).
2389 21. When testing a filter file using -bf, Exim was writing a message when it
2390 took the sender from a "From " line in the message, but it was not doing so
2391 when it took $return_path from a Return-Path: header line. It now does.
2393 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2394 with a valid header line field name (a series of printing characters
2395 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2397 23. Changed "disc" in the source to "disk" to conform to the documentation and
2398 the book and for uniformity.
2400 24. Ignore Sendmail's -Ooption=value command line item.
2402 25. When execve() failed while trying to run a command in a pipe transport,
2403 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2404 could be confused with a return value of 69 from the command itself. This
2405 has been changed to 127, the value the shell returns if it is asked to run
2406 a non-existent command. The wording for the related log line suggests a
2407 non-existent command as the problem.
2409 26. If received_header_text expands to an empty string, do not add a Received:
2410 header line to the message. (Well, it adds a token one on the spool, but
2411 marks it "old" so that it doesn't get used or transmitted.)
2413 27. Installed eximstats 1.28 (addition of -nt option).
2415 28. There was no check for failure on the call to getsockname() in the daemon
2416 code. This can fail if there is a shortage of resources on the system, with
2417 ENOMEM, for example. A temporary error is now given on failure.
2419 29. Contrary to the C standard, it seems that in some environments, the
2420 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2421 program. Exim now does this explicitly; it affects the formatting of
2422 timestamps using strftime().
2424 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2425 complaints. I've now initialized all the variables, to avoid this.
2427 32. Header lines added by a system filter were not being "seen" during
2428 transport-time rewrites.
2430 33. The info_callback() function passed to OpenSSL is set up with type void
2431 (*)(SSL *, int, int), as described somewhere. However, when calling the
2432 function (actually a macro) that sets it up, the type void(*)() is
2433 expected. I've put in a cast to prevent warnings from picky compilers.
2435 34. If a DNS black list lookup found a CNAME record, but there were no A
2436 records associated with the domain it pointed at, Exim crashed.
2438 35. If a DNS black list lookup returned more than one A record, Exim ignored
2439 all but the first. It now scans all returned addresses if a particular IP
2440 value is being sought. In this situation, the contents of the
2441 $dnslist_value variable are a list of all the addresses, separated by a
2444 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2445 to accept a host name and all its aliases if the forward lookup for any of
2446 them yielded the IP address of the incoming connection. Now it accepts only
2447 those names whose forward lookup yields the correct IP address. Any other
2448 names are discarded. This closes a loophole whereby a rogue DNS
2449 administrator could create reverse DNS records to break through a
2450 wildcarded host restriction in an ACL.
2452 37. If a user filter or a system filter that ran in a subprocess used any of
2453 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2454 the wrong values were passed to the pipe command ($thisaddress had the
2455 value of $0, $0 had the value of $1, etc). This bug was introduced by
2456 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2458 38. Improved the line breaking for long SMTP error messages from ACLs.
2459 Previously, if there was no break point between 40 and 75 characters, Exim
2460 left the rest of the message alone. Two changes have been made: (a) I've
2461 reduced the minimum length to 35 characters; (b) if it can't find a break
2462 point between 35 and 75 characters, it looks ahead and uses the first one
2463 that it finds. This may give the occasional overlong line, but at least the
2464 remaining text gets split now.
2466 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2467 file descriptors might be low, and that setting 1000 would always raise it.
2468 It turns out that in some environments, the limit is already over 1000 and
2469 that lowering it causes trouble. So now Exim takes care not to decrease it.
2471 40. When delivering a message, the value of $return_path is set to $sender_
2472 address at the start of routing (routers may change the value). By an
2473 oversight, this default was not being set up when an address was tested by
2474 -bt or -bv, which affected the outcome if any router or filter referred to
2477 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2478 log only when "message" or "log_message" are set. However, if one of the
2479 conditions was an address verification, or a call to a nested ACL, the
2480 messages generated by the underlying test were being passed through. This
2481 no longer happens. The underlying message is available in $acl_verify_
2482 message for both "message" and "log_message" expansions, so it can be
2483 passed through if needed.
2485 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2486 new expansion $bh_ to give the encoded byte string without charset
2487 translation. Translation happens only if iconv() is available; HAVE_ICONV
2488 indicates this at build time. HEADERS_CHARSET gives the charset to
2489 translate to; headers_charset can change it in the configuration, and
2490 "headers charset" can change it in an individual filter file.
2492 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2493 that creates RFC 2047 encoded "words" labels them as that charset instead
2494 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2495 expansion operator; (ii) when Exim creates a From: line for a local
2496 message; (iii) when a header line is rewritten to include a "phrase" part.
2498 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2499 buggy, causing it to skip the first lines of messages whose message ID
2500 ended in 'D'. This would not have bitten before Exim release 4.14, because
2501 message IDs were unlikely to end in 'D' before then. The effect was to have
2502 incorrect size information for certain domains.
2504 45. #include "config.h" was missing at the start of the crypt16.c module. This
2505 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2508 46. If there was a timeout during a "random" callout check, Exim treated it as
2509 a failure of the random address, and carried on sending RSET and the real
2510 address. If the delay was just some slowness somewhere, the response to the
2511 original RCPT would be taken as a response to RSET and so on, causing
2512 mayhem of various kinds.
2514 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2515 when I implemented it. It didn't allow for the fact that some option values
2516 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2517 the right test for positive values.
2519 48. Domain names in DNS records are case-independent. Exim always looks them up
2520 in lower case. Some resolvers return domain names in exactly the case they
2521 appear in the zone file, that is, they may contain uppercase letters. Not
2522 all resolvers do this - some return always lower case. Exim was treating a
2523 change of case by a resolver as a change of domain, similar to a widening
2524 of a domain abbreviation. This triggered its re-routing code and so it was
2525 trying to route what was effectively the same domain again. This normally
2526 caused routing to fail (because the router wouldn't handle the domain
2527 twice). Now Exim checks for this case specially, and just changes the
2528 casing of the domain that it ultimately uses when it transmits the message
2531 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2534 50. If a filter generated a file delivery with a non-absolute name (possible if
2535 no home directory exists for the router), the forbid_file option was not
2538 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2539 the existing equality matching.
2541 52. Exim was using ints instead of ino_t variables in some places where it was
2542 dealing with inode numbers.
2544 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2545 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2546 called TMPDIR, and if it finds it is different, it changes its value.
2548 54. The smtp_printf() function is now made available to local_scan() so
2549 additional output lines can be written before returning. There is also an
2550 smtp_fflush() function to enable the detection of a dropped connection.
2551 The variables smtp_input and smtp_batched_input are exported to
2554 55. Changed the default runtime configuration: the message "Unknown user"
2555 has been removed from the ACL, and instead placed on the localuser router,
2556 using the cannot_route_message feature. This means that any verification
2557 failures that generate their own messages won't get overridden. Similarly,
2558 the "Unrouteable address" message that was in the ACL for unverifiable
2559 relay addresses has also been removed.
2561 56. Added hosts_avoid_esmtp to the smtp transport.
2563 57. The exicyclog script was not checking for the esoteric option
2564 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2565 will work only if exicyclog is run under the appropriate euid.
2567 58. Following a discussion on the list, the rules by which Exim recognises line
2568 endings on incoming messages have been changed. The -dropcr and drop_cr
2569 options are now no-ops, retained only for backwards compatibility. The
2570 following line terminators are recognized: LF CRLF CR. However, special
2571 processing applies to CR:
2573 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2574 nor a local message in the state where . is a terminator.
2576 (ii) If a bare CR is encountered in a header line, an extra space is added
2577 after the line terminator so as not to end the header. The reasoning
2578 behind this is that bare CRs in header lines are most likely either
2579 to be mistakes, or people trying to play silly games.
2581 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2582 was being incorrectly given as 18 bytes larger than it should have been.
2583 This is a VOB (very old bug).
2585 60. This may never have affected anything current, but just in case it has:
2586 When the local host is found other than at the start of a list of hosts,
2587 the local host, those with the same MX, and any that follow, are discarded.
2588 When the list in question was part of a longer list of hosts, the following
2589 hosts (not currently being processed) were also being discarded. This no
2590 longer happens. I'm not sure if this situation could ever has previously
2593 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2596 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2597 header. This is something that is recommended in a new Internet Draft, and
2598 is something that is documented as being done by Sendmail. There are two
2599 possible values. For messages generated by the autoreply transport, Exim
2602 Auto-Submitted: auto-replied
2604 whereas for all other generated messages (e.g. bounces) it adds
2606 Auto-Submitted: auto-generated
2608 63. The "personal" condition in filters now includes a test for the
2609 Auto-Submitted: header. If it contains the string "auto-" the message it
2610 not considered personal.
2612 64. Added rcpt_include_affixes as a generic transport option.
2614 65. Added queue_only_override (default true).
2616 66. Added the syslog_duplication option.
2618 67. If what should have been the first header line of a message consisted of
2619 a space followed by a colon, Exim was mis-interpreting it as a header line.
2620 It isn't of course - it is syntactically invalid and should therefore be
2621 treated as the start of the message body. The misbehaviour could have
2622 caused a number of strange effects, including loss of data in subsequent
2623 header lines, and spool format errors.
2625 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2626 client host had authenticated. This control can now be exercised by an ACL
2627 for more flexibility.
2629 69. By default, callouts do not happen when testing with -bh. There is now a
2630 variant, -bhc, which does actually run the callout code, including
2631 consulting and updating the callout cache.
2633 70. Added support for saslauthd authentication, courtesy of Alexander
2636 71. If statvfs() failed on the spool or log directories while checking their
2637 size for availability, Exim confusingly gave the error "space shortage".
2638 Furthermore, in debugging mode it crashed with a floating point exception.
2639 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2640 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2641 serious problem, Exim now writes to the main and panic logs when this
2642 happens, with details of the failure. It then refuses to accept the
2643 incoming message, giving the message "spool directory problem" or "log
2644 directory problem" with a 421 code for SMTP messages.
2646 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2647 0, 1, and 2 exist, because some OS complain for execs without them (see
2648 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2649 descriptors. However, the code omitted to check that the open succeeded,
2650 causing mysterious errors if for some reason the permissions on /dev/null
2651 got screwed. Now Exim writes a message to the main and panic logs, and
2652 bombs out if it can't open /dev/null.
2654 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2655 interact so that it is all more flexible. It is supposed to remain
2656 backwards compatible. Also added extra_local_interfaces.
2658 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2659 to bomb out with an assertion failure - to the client this appears as a
2660 connection drop. This problem occurs in the part of the code that was taken
2661 from the Samba project. Fortunately, the assertion is in a very simple
2662 function, so I have fixed this by reproducing the function inline in the
2663 one place where it is called, and arranging for authentication to fail
2664 instead of killing the process with assert().
2666 75. The SPA client code was not working when the server requested OEM rather
2667 than Unicode encoding.
2669 76. Added code to make require_files with a specific uid setting more usable in
2670 the case where statting the file as root fails - usually a non-root-mounted
2671 NFS file system. When this happens and the failure is EACCES, Exim now
2672 forks a subprocess and does the per-uid checking as the relevant uid.
2674 77. Added process_log_path.
2676 78. If log_file_path was not explicitly set, a setting of check_log_space or
2677 check_log_inodes was ignored.
2679 79. If a space check for the spool or log partitions fails, the incident is now
2680 logged. Of course, in the latter case the data may get lost...
2682 80. Added the %p formatting code to string_format() so that it can be used to
2683 print addresses in debug_print(). Adjusted all the address printing in the
2684 debugging in store.c to use %p rather than %d.
2686 81. There was a concern that a line of code in smtp_in.c could overflow a
2687 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2688 initially expressed, the concern was not well-founded, because trailing
2689 spaces are removed early. However, if the trailing spaces were followed by
2690 a NULL, they did not get removed, so the overflow was possible. Two fixes
2693 (a) I re-wrote the offending code in a cleaner fashion.
2694 (b) If an incoming SMTP command contains a NULL character, it is rejected
2697 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2698 runs initgroups(), so that if the Exim user is in any additional groups,
2699 they will be used during message reception.
2705 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob