1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.109 2005/04/05 14:02:30 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" statement 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.
80 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
81 commented out examples of how to interface to a virus scanner and to
82 SpamAssassin. Also added commented examples of av_scanner and
83 spamd_address settings.
85 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
86 and controls are allowed in which ACLs. There were a couple of minor
87 errors. Some of the entries in the conditions table (which is a table of
88 where they are NOT allowed) were getting very unwieldy; rewrote them as a
89 negation of where the condition IS allowed.
91 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
93 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
94 header file does not have a version number, so I've had to invent a new
95 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
96 API. The code is untested by me (my Linux distribution still has 0.3.2 of
97 radiusclient), but it was contributed by a Radius user.
99 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
100 files or directories, for OpenSSL.
102 PH/19 When an Exim process that is running as root has to create an Exim log
103 file, it does so in a subprocess that runs as exim:exim so as to get the
104 ownership right at creation (otherwise, other Exim processes might see
105 the file with the wrong ownership). There was no test for failure of this
106 fork() call, which would lead to the process getting stuck as it waited
107 for a non-existent subprocess. Forks do occasionally fail when resources
108 run out. I reviewed all the other calls to fork(); they all seem to check
111 PH/20 When checking for unexpected SMTP input at connect time (before writing
112 the banner), Exim was not dealing correctly with a non-positive return
113 from the read() function. If the client had disconnected by this time,
114 the result was a log entry for a synchronization error with an empty
115 string after "input=" when read() returned zero. If read() returned -1
116 (an event I could not check), uninitialized data bytes were printed.
117 There were reports of junk text (parts of files, etc) appearing after
120 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
122 PH/22 Added support for macro redefinition, and (re)definition in between
123 driver and ACL definitions.
125 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
126 forgetting to use the resulting value; it was using the unexpanded value.
129 A note about Exim versions 4.44 and 4.50
130 ----------------------------------------
132 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
133 changes of various kinds. As a consequence, a big documentation update was
134 needed. This delayed the release for rather longer than seemed good, especially
135 in the light of a couple of (minor) security issues. Therefore, the changes
136 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
137 release. So 4.44 and 4.50 are in effect two different branches that both start
140 I have left the 4.50 change log unchanged; it contains all the changes since
141 4.43. The change log for 4.44 is below; many of its items are identical to
142 those for 4.50. This seems to be the most sensible way to preserve the
143 historical information.
149 1. Minor wording change to the doc/README.SIEVE file.
151 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
152 computation of the current number of files was incorrect.
154 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
155 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
158 4. Give more explanation in the error message when the command for a transport
159 filter fails to execute.
161 5. There are several places where Exim runs a non-Exim command in a
162 subprocess. The SIGUSR1 signal should be disabled for these processes. This
163 was being done only for the command run by the queryprogram router. It is
164 now done for all such subprocesses. The other cases are: ${run, transport
165 filters, and the commands run by the lmtp and pipe transports.
167 6. Added CONFIGURE_GROUP build-time option.
169 7. Some older OS have a limit of 256 on the maximum number of file
170 descriptors. Exim was using setrlimit() to set 1000 as a large value
171 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
172 systems. I've change it so that if it can't get 1000, it tries for 256.
174 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
175 was an oversight, and furthermore, ever since the addition of extra
176 controls (e.g. 4.43/32), the checks on when to allow different forms of
177 "control" were broken. There should now be diagnostics for all cases when a
178 control that does not make sense is encountered.
180 9. Added the /retain_sender option to "control=submission".
182 10. $recipients is now available in the predata ACL (oversight).
184 11. Tidy the search cache before the fork to do a delivery from a message
185 received from the command line. Otherwise the child will trigger a lookup
186 failure and thereby defer the delivery if it tries to use (for example) a
187 cached ldap connection that the parent has called unbind on.
189 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
190 of $address_data from the recipient verification was clobbered by the
193 13. The value of address_data from a sender verification is now available in
194 $sender_address_data in subsequent conditions in the ACL statement.
196 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
198 15. Added a new option "connect=<time>" to callout options, to set a different
201 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
202 was its contents. (It was OK if the option was not defined at all.)
204 17. A "Completed" log line is now written for messages that are removed from
205 the spool by the -Mrm option.
207 18. New variables $sender_verify_failure and $recipient_verify_failure contain
208 information about exactly what failed.
210 19. Added -dd to debug only the daemon process.
212 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
213 handles timeouts, both on the server side and network timeouts. Renamed the
214 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
216 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
217 It is now set to "smtps".
219 22. $host_address is now set to the target address during the checking of
222 23. When checking ignore_target_hosts for an ipliteral router, no host name was
223 being passed; this would have caused $sender_host_name to have been used if
224 matching the list had actually called for a host name (not very likely,
225 since this list is usually IP addresses). A host name is now passed as
228 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
229 code that specifies a non-restarting handler (typically sigaction() in
230 modern systems) in an attempt to fix a rare and obscure crash bug.
232 25. Narrowed the window for a race in the daemon that could cause it to ignore
233 SIGCHLD signals. This is not a major problem, because they are used only to
234 wake it up if nothing else does.
236 26. A malformed maildirsize file could cause Exim to calculate negative values
237 for the mailbox size or file count. Odd effects could occur as a result.
238 The maildirsize information is now recalculated if the size or filecount
241 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
242 support for a long time. Removed HAVE_SYS_VFS_H.
244 28. Installed the latest version of exipick from John Jetmore.
246 29. In an address list, if the pattern was not a regular expression, an empty
247 subject address (from a bounce message) matched only if the pattern was an
248 empty string. Non-empty patterns were not even tested. This was the wrong
249 because it is perfectly reasonable to use an empty address as part of a
250 database query. An empty address is now tested by patterns that are
251 lookups. However, all the other forms of pattern expect the subject to
252 contain a local part and a domain, and therefore, for them, an empty
253 address still always fails if the pattern is not itself empty.
255 30. Exim went into a mad DNS loop when attempting to do a callout where the
256 host was specified on an smtp transport, and looking it up yielded more
259 31. Re-factored the code for checking spool and log partition space into a
260 function that finds that data and another that does the check. The former
261 is then used to implement four new variables: $spool_space, $log_space,
262 $spool_inodes, and $log_inodes.
264 32. The RFC2047 encoding function was originally intended for short strings
265 such as real names; it was not keeping to the 75-character limit for
266 encoded words that the RFC imposes. It now respects the limit, and
267 generates multiple encoded words if necessary. To be on the safe side, I
268 have increased the buffer size for the ${rfc2047: expansion operator from
271 33. It is now permitted to omit both strings after an "if" condition; if the
272 condition is true, the result is "true". As before, when the second string
273 is omitted, a false condition yields an empty string. This makes it less
274 cumbersome to write custom ACL and router conditions.
276 34. Failure to deliver a bounce message always caused it to be frozen, even if
277 there was an errors_to setting on the router. The errors_to setting is now
280 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
281 canonical form (fully expanded) before being placed in
282 $sender_host_address.
284 36. The table in the code that translates DNS record types into text (T_A to
285 "A" for instance) was missing entries for NS and CNAME. It is just possible
286 that this could have caused confusion if both these types were looked up
287 for the same domain, because the text type is used as part of Exim's
288 per-process caching. But the chance of anyone hitting this buglet seems
291 37. The dnsdb lookup has been extended in a number of ways.
293 (1) There is a new type, "zns", which walks up the domain tree until it
294 finds some nameserver records. It should be used with care.
296 (2) There is a new type, "mxh", which is like "mx" except that it returns
297 just the host names, not the priorities.
299 (3) It is now possible to give a list of domains (or IP addresses) to be
300 looked up. The behaviour when one of the lookups defers can be
301 controlled by a keyword.
303 (4) It is now possible to specify the separator character for use when
304 multiple records are returned.
306 38. The dnslists ACL condition has been extended: it is now possible to supply
307 a list of IP addresses and/or domains to be looked up in a particular DNS
310 39. Added log_selector=+queue_time_overall.
312 40. When running the queue in the test harness, wait just a tad after forking a
313 delivery process, to get repeatability of debugging output.
315 41. Include certificate and key file names in error message when GnuTLS fails
316 to set them up, because the GnuTLS error message doesn't include the name
317 of the failing file when there is a problem reading it.
319 42. Allow both -bf and -bF in the same test run.
321 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
323 44. The "Exiscan patch" is now merged into the mainline Exim source.
325 45. Sometimes the final signoff response after QUIT could fail to get
326 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
327 < 0 before doing a fflush(). This bug looks as though it goes back to the
328 introduction of TLS in release 3.20, but "sometimes" must have been rare
329 because the tests only now provoked it.
331 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
332 (this can affect the format of dates).
334 47. exim_tidydb, when checking for the continued existence of a message for
335 which it has found a message-specific retry record, was not finding
336 messages that were in split spool directories. Consequently, it was
337 deleting retry records that should have stayed in existence.
339 48. Steve fixed some bugs in eximstats.
341 49. The SPA authentication driver was not abandoning authentication and moving
342 on to the next authenticator when an expansion was forced to fail,
343 contradicting the general specification for all authenticators. Instead it
344 was generating a temporary error. It now behaves as specified.
346 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
347 (the order specifies the preference for clients). The order is now AES256,
348 AES128, 3DES, ARCFOUR128.
350 51. Small patch to Sieve code - explicitly set From: when generating an
353 52. Exim crashed if a remote delivery caused a very long error message to be
354 recorded - for instance if somebody sent an entire SpamAssassin report back
355 as a large number of 550 error lines. This bug was coincidentally fixed by
356 increasing the size of one of Exim's internal buffers (big_buffer) that
357 happened as part of the Exiscan merge. However, to be on the safe side, I
358 have made the code more robust (and fixed the comments that describe what
361 53. Now that there can be additional text after "Completed" in log lines (if
362 the queue_time_overall log selector is set), a one-byte patch to exigrep
363 was needed to allow it to recognize "Completed" as not the last thing in
366 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
367 patch that reportedly fixes this has been added. I am not expert enough to
368 create a test for it. This is what the patch creator wrote:
370 "I found a little strange behaviour of ldap code when working with
371 Windows 2003 AD Domain, where users was placed in more than one
372 Organization Units. When I tried to give exim partial DN, the exit code
373 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
374 But simultaneously result of request was absolutely normal ldap result,
375 so I produce this patch..."
377 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
378 REFERENCE, so I have modified the code to exclude the patch when that macro
381 55. Some experimental protocols are using DNS PTR records for new purposes. The
382 keys for these records are domain names, not reversed IP addresses. The
383 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
384 leaves it alone. Component reversal etc. now happens only for IP addresses.
387 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
389 57. Double the size of the debug message buffer (to 2048) so that more of very
390 long debug lines gets shown.
392 58. The exicyclog utility now does better if the number of log files to keep
393 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
395 59. Two changes related to the smtp_active_hostname option:
397 (1) $smtp_active_hostname is now available as a variable.
398 (2) The default for smtp_banner uses $smtp_active_hostname instead
399 of $primary_hostname.
401 60. The host_aton() function is supposed to be passed a string that is known
402 to be a valid IP address. However, in the case of IPv6 addresses, it was
403 not checking this. This is a hostage to fortune. Exim now panics and dies
404 if the condition is not met. A case was found where this could be provoked
405 from a dnsdb PTR lookup with an IPv6 address that had more than 8
406 components; fortuitously, this particular loophole had already been fixed
407 by change 4.50/55 above.
409 If there are any other similar loopholes, the new check in host_aton()
410 itself should stop them being exploited. The report I received stated that
411 data on the command line could provoke the exploit when Exim was running as
412 exim, but did not say which command line option was involved. All I could
413 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
417 61. There was a buffer overflow vulnerability in the SPA authentication code
418 (which came originally from the Samba project). I have added a test to the
419 spa_base64_to_bits() function which I hope fixes it.
422 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
423 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
425 63. The daemon start-up calls getloadavg() while still root for those OS that
426 need the first call to be done as root, but it missed one case: when
427 deliver_queue_load_max is set with deliver_drop_privilege. This is
428 necessary for the benefit of the queue runner, because there is no re-exec
429 when deliver_drop_privilege is set.
431 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
434 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
435 was in use, was not putting the data itself into the right store pool;
436 consequently, it could be overwritten for a subsequent message in the same
437 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
440 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
442 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
443 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
444 the function were treating the return as a boolean value, which happened to
445 work because 0=false and not-0=true, but is not correct code.
447 68. The host_aton() function was not handling scoped IPv6 addresses (those
448 with, for example, "%eth0" on the end) correctly.
450 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
451 negated items (that is, ~something) in unsigned ints. Some compilers
452 apparently mutter when there is no cast.
454 70. If an address verification called from an ACL failed, and did not produce a
455 user-specific message (i.e. there was only a "system" message), nothing was
456 put in $acl_verify_message. In this situation, it now puts the system
459 71. Change 4.23/11 added synchronization checking at the start of an SMTP
460 session; change 4.31/43 added the unwanted input to the log line - except
461 that it did not do this in the start of session case. It now does.
463 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
464 This is wrong and can cause the other end to generate a synchronization
465 error if it is another Exim or anything else that does the synchronization
466 check. A QUIT command is no longer sent after a timeout.
468 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
471 74. The defer_ok option of callout verification was not working if it was used
472 when verifying addresses in header lines, that is, for this case:
474 verify = header_sender/callout=defer_ok
476 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
477 those file descriptors could be used for SMTP connections. If anything
478 wrote to stderr (the example that came up was "warn" in embedded Perl), it
479 could be sent to the SMTP client, causing chaos. The daemon now opens
480 stdin, stdout, and stderr to /dev/null when it puts itself into the
483 76. Arrange for output from Perl's "warn" command to be written to Exim's main
484 log by default. The user can override this with suitable Perl magic.
486 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
487 the log message when discard triggers, was not working for the DATA ACL or
488 for the non-SMTP ACL.
490 78. Error message wording change in sieve.c.
492 79. If smtp_accept_max_per_host was set, the number of connections could be
493 restricted to fewer than expected, because the daemon was trying to set up
494 a new connection before checking whether the processes handling previous
495 connections had finished. The check for completed processes is now done
496 earlier. On busy systems, this bug wouldn't be noticed because something
497 else would have woken the daemon, and it would have reaped the completed
500 80. If a message was submitted locally by a user whose login name contained one
501 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
502 It caused a spool format error. I have fixed the spool reading code. A
503 related problem was that the "from" clause in the Received: line became
504 illegal because of the space(s). It is now covered by ${quote_local_part.
506 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
509 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
511 83. Patch from Sieve maintainer to fix unterminated string problem in
514 84. Some minor changes to the Linux configuration files to help with other
515 OS variants using glibc.
517 85. One more patch for Sieve to update vacation handling to latest spec.
520 ----------------------------------------------------
521 See the note above about the 4.44 and 4.50 releases.
522 ----------------------------------------------------
528 1. Change 4.43/35 introduced a bug that caused file counts to be
529 incorrectly computed when quota_filecount was set in an appendfile
532 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
533 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
536 3. Give more explanation in the error message when the command for a transport
537 filter fails to execute.
539 4. There are several places where Exim runs a non-Exim command in a
540 subprocess. The SIGUSR1 signal should be disabled for these processes. This
541 was being done only for the command run by the queryprogram router. It is
542 now done for all such subprocesses. The other cases are: ${run, transport
543 filters, and the commands run by the lmtp and pipe transports.
545 5. Some older OS have a limit of 256 on the maximum number of file
546 descriptors. Exim was using setrlimit() to set 1000 as a large value
547 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
548 systems. I've change it so that if it can't get 1000, it tries for 256.
550 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
551 was an oversight, and furthermore, ever since the addition of extra
552 controls (e.g. 4.43/32), the checks on when to allow different forms of
553 "control" were broken. There should now be diagnostics for all cases when a
554 control that does not make sense is encountered.
556 7. $recipients is now available in the predata ACL (oversight).
558 8. Tidy the search cache before the fork to do a delivery from a message
559 received from the command line. Otherwise the child will trigger a lookup
560 failure and thereby defer the delivery if it tries to use (for example) a
561 cached ldap connection that the parent has called unbind on.
563 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
564 of $address_data from the recipient verification was clobbered by the
567 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
568 was its contents. (It was OK if the option was not defined at all.)
570 11. A "Completed" log line is now written for messages that are removed from
571 the spool by the -Mrm option.
573 12. $host_address is now set to the target address during the checking of
576 13. When checking ignore_target_hosts for an ipliteral router, no host name was
577 being passed; this would have caused $sender_host_name to have been used if
578 matching the list had actually called for a host name (not very likely,
579 since this list is usually IP addresses). A host name is now passed as
582 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
583 code that specifies a non-restarting handler (typically sigaction() in
584 modern systems) in an attempt to fix a rare and obscure crash bug.
586 15. Narrowed the window for a race in the daemon that could cause it to ignore
587 SIGCHLD signals. This is not a major problem, because they are used only to
588 wake it up if nothing else does.
590 16. A malformed maildirsize file could cause Exim to calculate negative values
591 for the mailbox size or file count. Odd effects could occur as a result.
592 The maildirsize information is now recalculated if the size or filecount
595 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
596 support for a long time. Removed HAVE_SYS_VFS_H.
598 18. Updated exipick to current release from John Jetmore.
600 19. Allow an empty sender to be matched against a lookup in an address list.
601 Previously the only cases considered were a regular expression, or an
604 20. Exim went into a mad DNS lookup loop when doing a callout where the
605 host was specified on the transport, if the DNS lookup yielded more than
608 21. The RFC2047 encoding function was originally intended for short strings
609 such as real names; it was not keeping to the 75-character limit for
610 encoded words that the RFC imposes. It now respects the limit, and
611 generates multiple encoded words if necessary. To be on the safe side, I
612 have increased the buffer size for the ${rfc2047: expansion operator from
615 22. Failure to deliver a bounce message always caused it to be frozen, even if
616 there was an errors_to setting on the router. The errors_to setting is now
619 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
620 canonical form (fully expanded) before being placed in
621 $sender_host_address.
623 24. Updated eximstats to version 1.33
625 25. Include certificate and key file names in error message when GnuTLS fails
626 to set them up, because the GnuTLS error message doesn't include the name
627 of the failing file when there is a problem reading it.
629 26. Expand error message when OpenSSL has problems setting up cert/key files.
632 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
633 (this can affect the format of dates).
635 28. exim_tidydb, when checking for the continued existence of a message for
636 which it has found a message-specific retry record, was not finding
637 messages that were in split spool directories. Consequently, it was
638 deleting retry records that should have stayed in existence.
640 29. eximstats updated to version 1.35
641 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
642 1.35 - bugfix such that pie charts by volume are generated correctly
644 30. The SPA authentication driver was not abandoning authentication and moving
645 on to the next authenticator when an expansion was forced to fail,
646 contradicting the general specification for all authenticators. Instead it
647 was generating a temporary error. It now behaves as specified.
649 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
650 (the order specifies the preference for clients). The order is now AES256,
651 AES128, 3DES, ARCFOUR128.
653 31. Small patch to Sieve code - explicitly set From: when generating an
656 32. Exim crashed if a remote delivery caused a very long error message to be
657 recorded - for instance if somebody sent an entire SpamAssassin report back
658 as a large number of 550 error lines. This bug was coincidentally fixed by
659 increasing the size of one of Exim's internal buffers (big_buffer) that
660 happened as part of the Exiscan merge. However, to be on the safe side, I
661 have made the code more robust (and fixed the comments that describe what
664 33. Some experimental protocols are using DNS PTR records for new purposes. The
665 keys for these records are domain names, not reversed IP addresses. The
666 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
667 leaves it alone. Component reversal etc. now happens only for IP addresses.
670 34. The host_aton() function is supposed to be passed a string that is known
671 to be a valid IP address. However, in the case of IPv6 addresses, it was
672 not checking this. This is a hostage to fortune. Exim now panics and dies
673 if the condition is not met. A case was found where this could be provoked
674 from a dnsdb PTR lookup with an IPv6 address that had more than 8
675 components; fortuitously, this particular loophole had already been fixed
676 by change 4.50/55 or 4.44/33 above.
678 If there are any other similar loopholes, the new check in host_aton()
679 itself should stop them being exploited. The report I received stated that
680 data on the command line could provoke the exploit when Exim was running as
681 exim, but did not say which command line option was involved. All I could
682 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
686 35. There was a buffer overflow vulnerability in the SPA authentication code
687 (which came originally from the Samba project). I have added a test to the
688 spa_base64_to_bits() function which I hope fixes it.
691 36. The daemon start-up calls getloadavg() while still root for those OS that
692 need the first call to be done as root, but it missed one case: when
693 deliver_queue_load_max is set with deliver_drop_privilege. This is
694 necessary for the benefit of the queue runner, because there is no re-exec
695 when deliver_drop_privilege is set.
697 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
698 was in use, was not putting the data itself into the right store pool;
699 consequently, it could be overwritten for a subsequent message in the same
700 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
703 38. Sometimes the final signoff response after QUIT could fail to get
704 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
705 < 0 before doing a fflush(). This bug looks as though it goes back to the
706 introduction of TLS in release 3.20, but "sometimes" must have been rare
707 because the tests only now provoked it.
713 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
714 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
715 Now it returns an integer. A number of calls were still expecting a T/F
716 return. Fortuitously, in all cases, the tests worked in OK situations,
717 which is the norm. However, things would have gone wrong on any write
718 failures on the smtp file descriptor. This function is used when sending
719 messages over SMTP and also when doing verify callouts.
721 2. When Exim is called to do synchronous delivery of a locally submitted
722 message (the -odf or -odi options), it no longer closes stderr before doing
725 3. Implemented the mua_wrapper option.
727 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
729 5. Implemented the functions header_remove(), header_testname(),
730 header_add_at_position(), and receive_remove_recipient(), and exported them
733 6. If an ACL "warn" statement specified the addition of headers, Exim already
734 inserted X-ACL-Warn: at the start if there was no header name. However, it
735 was not making this test for the second and subsequent header lines if
736 there were newlines in the string. This meant that an invalid header could
737 be inserted if Exim was badly configured.
739 7. Allow an ACL "warn" statement to add header lines at the start or after all
740 the Received: headers, as well as at the end.
742 8. Added the rcpt_4xx retry error code.
744 9. Added postmaster_mailfrom=xxx to callout verification option.
746 10. Added mailfrom=xxxx to the callout verification option, for verify=
749 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
750 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
752 12. Inserted some casts to stop certain compilers warning when using pointer
753 differences as field lengths or precisions in printf-type calls (mostly
754 affecting debugging statements).
756 13. Added optional readline() support for -be (dynamically loaded).
758 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
759 same clock tick as a message's arrival, so that its received time was the
760 same as the "first fail" time on the retry record, and that message
761 remained on the queue past the ultimate address timeout, every queue runner
762 would try a delivery (because it was past the ultimate address timeout) but
763 after another failure, the ultimate address timeout, which should have then
764 bounced the address, did not kick in. This was a "< instead of <=" error;
765 in most cases the first failure would have been in the next clock tick
766 after the received time, and all would be well.
768 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
769 being recognized when the domain list was tested by the match_domain
770 condition in an expansion string.
772 16. Added the ${str2b64: operator.
774 17. Exim was always calling setrlimit() to set a large limit for the number of
775 processes, without checking whether the existing limit was already
776 adequate. (It did check for the limit on file descriptors.) Furthermore,
777 errors from getrlimit() and setrlimit() were being ignored. Now they are
778 logged to the main and panic logs, but Exim does carry on, to try to do its
779 job under whatever limits there are.
781 18. Imported PCRE 5.0.
783 19. Trivial typo in log message " temporarily refused connection" (the leading
786 20. If the log selector return_path_on_delivery was set and an address was
787 redirected to /dev/null, the delivery process crashed because it assumed
788 that a return path would always be set for a "successful" delivery. In this
789 case, the whole delivery is bypassed as an optimization, and therefore no
792 21. Internal re-arrangement: the function for sending a challenge and reading
793 a response while authentication was assuming a zero-terminated challenge
794 string. It's now changed to take a pointer and a length, to allow for
795 binary data in such strings.
797 22. Added the cyrus_sasl authenticator (code supplied by MBM).
799 23. Exim was not respecting finduser_retries when seeking the login of the
800 uid under which it was called; it was always trying 10 times. (The default
801 setting of finduser_retries is zero.) Also, it was sleeping after the final
802 failure, which is pointless.
804 24. Implemented tls_on_connect_ports.
806 25. Implemented acl_smtp_predata.
808 26. If the domain in control=submission is set empty, Exim assumes that the
809 authenticated id is a complete email address when it generates From: or
810 Sender: header lines.
812 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
813 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
814 chown and chgrp in /bin and hostname in /usr/bin.
816 28. Exim was keeping the "process log" file open after each use, just as it
817 does for the main log. This opens the possibility of it remaining open for
818 long periods when the USR1 signal hits a daemon. Occasional processlog
819 errors were reported, that could have been caused by this. Anyway, it seems
820 much more sensible not to leave this file open at all, so that is what now
823 29. The long-running daemon process does not normally write to the log once it
824 has entered its main loop, and it closes the log before doing so. This is
825 so that log files can straightforwardly be renamed and moved. However,
826 there are a couple of unusual error situations where the daemon does write
827 log entries, and I had neglected to close the log afterwards.
829 30. The text of an SMTP error response that was received during a remote
830 delivery was being truncated at 512 bytes. This is too short for some of
831 the long messages that one sometimes sees. I've increased the limit to
834 31. It is now possible to make retry rules that apply only when a message has a
835 specific sender, in particular, an empty sender.
837 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
838 it possible to be selective about when SMTP synchronization is enforced.
840 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
842 32. Implemented hosts_connection_nolog.
844 33. Added an ACL for QUIT.
846 34. Setting "delay_warning=" to disable warnings was not working; it gave a
849 35. Added mailbox_size and mailbox_filecount to appendfile.
851 36. Added control = no_multiline_responses to ACLs.
853 37. There was a bug in the logic of the code that waits for the clock to tick
854 in the case where the clock went backwards by a substantial amount such
855 that the microsecond fraction of "now" was more than the microsecond
856 fraction of "then" (but the whole seconds number was less).
858 38. Added support for the libradius Radius client library this is found on
859 FreeBSD (previously only the radiusclient library was supported).
865 1. When certain lookups returned multiple values in the form name=value, the
866 quoting of the values was not always being done properly. Specifically:
867 (a) If the value started with a double quote, but contained no whitespace,
869 (b) If the value contained whitespace other than a space character (i.e.
870 tabs or newlines or carriage returns) it was not quoted.
871 This fix has been applied to the mysql and pgsql lookups by writing a
872 separate quoting function and calling it from the lookup code. The fix
873 should probably also be applied to nisplus, ibase and oracle lookups, but
874 since I cannot test any of those, I have not disturbed their existing code.
876 2. A hit in the callout cache for a specific address caused a log line with no
877 reason for rejecting RCPT. Now it says "Previous (cached) callout
878 verification failure".
880 3. There was an off-by-one bug in the queryprogram router. An over-long
881 return line was truncated at 256 instead of 255 characters, thereby
882 overflowing its buffer with the terminating zero. As well as fixing this, I
883 have increased the buffer size to 1024 (and made a note to document this).
885 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
886 when Exim is waiting for an SMTP response from a remote server, Exim
887 restarts its select() call on the socket, thereby resetting its timeout.
888 This is not a problem when such interrupts are rare. Somebody set up a cron
889 job to run exiwhat every 2 minutes, which is less than the normal select()
890 timeout (5 or 10 minutes). This meant that the select() timeout never
891 kicked in because it was always reset. I have fixed this by comparing the
892 time when an interrupt arrives with the time at the start of the first call
893 to select(). If more time than the timeout has elapsed, the interrupt is
894 treated as a timeout.
896 5. Some internal re-factoring in preparation for the addition of Sieve
897 extensions (by MH). In particular, the "personal" test is moved to a
898 separate function, and given an option for scanning Cc: and Bcc: (which is
899 not set for Exim filters).
901 6. When Exim created an email address using the login of the caller as the
902 local part (e.g. when creating a From: or Sender: header line), it was not
903 quoting the local part when it contained special characters such as @.
905 7. Installed new OpenBSD configuration files.
907 8. Reworded some messages for syntax errors in "and" and "or" conditions to
908 try to make them clearer.
910 9. Callout options, other than the timeout value, were being ignored when
911 verifying sender addresses in header lines. For example, when using
913 verify = header_sender/callout=no_cache
915 the cache was (incorrectly) being used.
917 10. Added a missing instance of ${EXE} to the exim_install script; this affects
918 only the Cygwin environment.
920 11. When return_path_on_delivery was set as a log selector, if different remote
921 addresses in the same message used different return paths and parallel
922 remote delivery occurred, the wrong values would sometimes be logged.
923 (Whenever a remote delivery process finished, the return path value from
924 the most recently started remote delivery process was logged.)
926 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
927 lines when AUTH and/or TLS are in use. This is the "received protocol"
928 field. Exim used to use "asmtp" for authenticated SMTP, without any
929 indication (in the protocol name) for TLS use. Now it follows the RFC and
930 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
931 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
932 names appear in log lines as well as in Received: header lines.
934 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
935 extensions, and comparison tests, and to fix some bugs.
937 14. Changes to the "personal" filter test:
939 (1) The test was buggy in that it was just doing the equivalent of
940 "contains" tests on header lines. For example, if a user's address was
941 anne@some.where, the "personal" test would incorrectly be true for
943 To: susanne@some.where
945 This test is now done by extracting each address from the header in turn,
946 and checking the entire address. Other tests that are part of "personal"
947 are now done using regular expressions (for example, to check local parts
948 of addresses in From: header lines).
950 (2) The list of non-personal local parts in From: addresses has been
951 extended to include "listserv", "majordomo", "*-request", and "owner-*",
952 taken from the Sieve specification recommendations.
954 (3) If the message contains any header line starting with "List-" it is
955 treated as non-personal.
957 (4) The test for "circular" in the Subject: header line has been removed
958 because it now seems ill-conceived.
960 15. Minor typos in src/EDITME comments corrected.
962 16. Installed latest exipick from John Jetmore.
964 17. If headers_add on a router specified a text string that was too long for
965 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
966 of string_sprintf() is now avoided.
968 18. $message_body_size was not set (it was always zero) when running the DATA
969 ACL and the local_scan() function.
971 19. For the "mail" command in an Exim filter, no default was being set for
972 the once_repeat time, causing a random time value to be used if "once" was
973 specified. (If the value happened to be <= 0, no repeat happened.) The
974 default is now 0s, meaning "never repeat". The "vacation" command was OK
975 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
976 (I found it when inspecting the code).
978 20. There is now an overall timeout for performing a callout verification. It
979 defaults to 4 times the callout timeout, which applies to individual SMTP
980 commands during the callout. The overall timeout applies when there is more
981 than one host that can be tried. The timeout is checked before trying the
982 next host. This prevents very long delays if there are a large number of
983 hosts and all are timing out (e.g. when the network connections are timing
984 out). The value of the overall timeout can be changed by specifying an
985 additional sub-option for "callout", called "maxwait". For example:
987 verify = sender/callout=5s,maxwait=20s
989 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
990 to the end before writing, but this should make it even safer).
992 22. Exim was forgetting that it had advertised PIPELINING for the second and
993 subsequent messages on an SMTP connection. It was also not resetting its
994 memory on STARTTLS and an internal HELO.
996 23. When Exim logs an SMTP synchronization error within a session, it now
997 records whether PIPELINING has been advertised or not.
999 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1000 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1001 rather than long int.
1003 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1005 26. Added the never_mail option to autoreply.
1011 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1012 crash if the getsockname() call failed; this can happen if a connection is
1013 closed very soon after it is established. The problem was simply in the
1014 order in which certain operations were done, causing Exim to try to write
1015 to the SMTP stream before it had set up the file descriptor. The bug has
1016 been fixed by making things happen in the correct order.
1022 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1023 before the connection was closed, thus losing the rejection response.
1025 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1026 some early Solaris releases, but causes trouble in current releases where
1027 socklen_t is defined.
1029 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1032 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1035 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1036 that releasing the top of it at the end releases what was used for sub-
1037 expansions (unless the block got too big). However, discard this block if
1038 the first thing is a variable or header, so that we can use its block when
1039 it is dynamic (useful for very large $message_headers, for example).
1041 6. Lookups now cache *every* query, not just the most recent. A new, separate
1042 store pool is used for this. It can be recovered when all lookup caches are
1043 flushed. Lookups now release memory at the end of their result strings.
1044 This has involved some general refactoring of the lookup sources.
1046 7. Some code has been added to the store_xxx() functions to reduce the amount
1047 of flapping under certain conditions.
1049 8. log_incoming_interface used to affect only the <= reception log lines. Now
1050 it causes the local interface and port to be added to several more SMTP log
1051 lines, for example "SMTP connection from", and rejection lines.
1053 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1055 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1057 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1058 could be overwritten at the end of the current message (or the start of a
1059 new message if it was set in a HELO ACL). The value is now preserved for
1060 the duration of the SMTP connection.
1062 12. If a transport had a headers_rewrite setting, and a matching header line
1063 contained an unqualified address, that address was qualified, even if it
1064 did not match any rewriting rules. The underlying bug was that the values
1065 of the flags that permit the existence of unqualified sender and recipient
1066 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1067 non-local messages, and by -bnq for local messages) were not being
1068 preserved with the message after it was received.
1070 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1071 "next input=" as part of the text comprising the host identity instead of
1072 the correct text. The code was using the same buffer for two different
1073 strings. However, depending on which order the printing function evaluated
1074 its arguments, the bug did not always show up. Under Linux, for example, my
1075 test suite worked just fine.
1077 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1078 doesn't work with some older versions of Perl. It has been changed to "my",
1079 which in any case is probably the better facility to use.
1081 15. A really picky compiler found some instances of statements for creating
1082 error messages that either had too many or two few arguments for the format
1085 16. The size of the buffer for calls to the DNS resolver has been increased
1086 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1087 for addresses that have a lot of PTR records. This alleviates a problem; it
1088 does not fully solve it.
1090 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1091 buffer now truncates the list and logs the incident, which is the same
1092 action as happens when Exim is looking up a host name and its aliases.
1093 Previously in this situation something unpredictable would happen;
1094 sometimes it was "internal error: store_reset failed".
1096 18. If a server dropped the connection unexpectedly when an Exim client was
1097 using GnuTLS and trying to read a response, the client delivery process
1098 crashed while trying to generate an error log message.
1100 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1101 string, for example:
1103 warn message = H1: something\nH2: something
1105 the text was added as a single header line from Exim's point of view
1106 though it ended up OK in the delivered message. However, searching for the
1107 second and subsequent header lines using $h_h2: did not work. This has been
1108 fixed. Similarly, if a system filter added multiple headers in this way,
1109 the routers could not see them.
1111 20. Expanded the error message when iplsearch is called with an invalid key to
1112 suggest using net-iplsearch in a host list.
1114 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1115 ACLs are no longer actually imposed (and a message to that effect is
1118 22. If a "gecos" field in a passwd entry contained escaped characters, in
1119 particular, if it contained a \" sequence, Exim got it wrong when building
1120 a From: or a Sender: header from that name. A second bug also caused
1121 incorrect handling when an unquoted " was present following a character
1122 that needed quoting.
1124 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1125 was not being matched caselessly.
1127 24. Arranged for all hyphens in the exim.8 source to be escaped with
1130 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1131 properly. Recipient callout cache records were still being keyed to include
1132 the sender, even when use_sender was set false. This led to far more
1133 callouts that were necessary. The sender is no longer included in the key
1134 when use_sender is false.
1136 26. Added "control = submission" modifier to ACLs.
1138 27. Added the ${base62d: operator to decode base 62 numbers.
1140 28. dnsdb lookups can now access SRV records.
1142 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1143 the configuration file.
1145 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1146 (-v) mode. This makes the output for a verbose queue run more intelligible.
1148 31. Added a use_postmaster feature to recipient callouts.
1150 32. Added the $body_zerocount variable, containing the number of binary zero
1151 bytes in the message body.
1153 33. The time of last modification of the "new" subdirectory is now used as the
1154 "mailbox time last read" when there is a quota error for a maildir
1157 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1159 35. Added +ignore_unknown as a special item in host lists.
1161 36. Code for decoding IPv6 addresses in host lists is now included, even if
1162 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1163 address was recognized as an IP address, but was then not correctly decoded
1164 into binary, causing unexpected and incorrect effects when compared with
1171 1. Very minor rewording of debugging text in manualroute to say "list of
1172 hosts" instead of "hostlist".
1174 2. If verify=header_syntax was set, and a header line with an unqualified
1175 address (no domain) and a large number of spaces between the end of the
1176 name and the colon was received, the reception process suffered a buffer
1177 overflow, and (when I tested it) crashed. This was caused by some obsolete
1178 code that should have been removed. The fix is to remove it!
1180 3. When running in the test harness, delay a bit after writing a bounce
1181 message to get a bit more predictability in the log output.
1183 4. Added a call to search_tidyup() just before forking a reception process. In
1184 theory, someone could use a lookup in the expansion of smtp_accept_max_
1185 per_host which, without the tidyup, could leave open a database connection.
1187 5. Added the variables $recipient_data and $sender_data which get set from a
1188 lookup success in an ACL "recipients" or "senders" condition, or a router
1189 "senders" option, similar to $domain_data and $local_part_data.
1191 6. Moved the writing of debug_print from before to after the "senders" test
1194 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1195 problems for message scanning, either using a data ACL, or using
1196 local_scan() because the Received: header was not generated till after they
1197 were called (in order to set the time as the time of reception completion).
1198 I have revised the way this works. The header is now generated after the
1199 body is received, but before the ACL or local_scan() are called. After they
1200 are run, the timestamp in the header is updated.
1206 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1207 before starting a queue runner without re-exec. This happened only when
1208 deliver_drop_privilege was set or when the Exim user was set to root. The
1209 effect of the bug was that timeouts during subsequent deliveries caused
1210 crashes instead of being properly handled. The handler is now left at its
1211 default (and expected) setting.
1213 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1214 message, again when deliver_drop_privilege is set or Exim is run as root.
1215 The bug described in (1) was not present in this case, but the tidying up
1216 of the other signals was missing. I have made the two cases consistent.
1218 3. The ignore_target_hosts setting on a manualroute router was being ignored
1219 for hosts that were looked up using the /MX notation.
1221 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1224 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1225 operated on the sender address. After changing the $sender_address to <>
1226 for the sender address verify, Exim was re-instated it as the original
1227 (before rewriting) address, but remembering that it had rewritten it, so it
1228 wasn't rewriting it again. This bug also had the effect of breaking the
1229 sender address verification caching when the sender address was rewritten.
1231 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1232 This has been changed so that if the ip literal address matches
1233 ignore_target_hosts, the router declines.
1235 7. Added expansion conditions match_domain, match_address, and match_local_
1236 part (NOT match_host).
1238 8. The placeholder for the Received: header didn't have a length field set.
1240 9. Added code to Exim itself and to exim_lock to test for a specific race
1241 condition that could lead to file corruption when using MBX delivery. The
1242 issue is with the lockfile that is created in /tmp. If this file is removed
1243 after a process has opened it but before that process has acquired a lock,
1244 there is the potential for a second process to recreate the file and also
1245 acquire a lock. This could lead to two Exim processes writing to the file
1246 at the same time. The added code performs the same test as UW imapd; it
1247 checks after acquiring the lock that its file descriptor still refers to
1248 the same named file.
1250 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1251 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1254 11. Added the smtp_active_hostname option. If used, this will typically be made
1255 to depend on the incoming interface address. Because $interface_address is
1256 not set up until the daemon has forked a reception process, error responses
1257 that can happen earlier (such as "too many connections") no longer contain
1260 12. If an expansion in a condition on a "warn" statement fails because a lookup
1261 defers, the "warn" statement is abandoned, and the next ACL statement is
1262 processed. Previously this caused the whole ACL to be aborted.
1264 13. Added the iplsearch lookup type.
1266 14. Added ident_timeout as a log selector.
1268 15. Added tls_certificate_verified as a log selector.
1270 16. Added a global option tls_require_ciphers (compare the smtp transport
1271 option of the same name). This controls incoming TLS connections.
1273 17. I finally figured out how to make tls_require_ciphers do a similar thing
1274 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1275 before starting the TLS session.
1277 18. Tabs are now shown as \t in -bP output.
1279 19. If the log selector return_path_on_delivery was set, Exim crashed when
1280 bouncing a message because it had too many Received: header lines.
1282 20. If two routers both had headers_remove settings, and the first one included
1283 a superfluous trailing colon, the final name in the first list and the
1284 first name in the second list were incorrectly joined into one item (with a
1285 colon in the middle).
1291 1. Added -C and -D options to the exinext utility, mainly to make it easier
1292 to include in the automated testing, but these could be helpful when
1293 multiple configurations are in use.
1295 2. The exinext utility was not formatting the output nicely when there was
1296 an alternate port involved in the retry record key, nor when there was a
1297 message id as well (for retries that were specific to a specific message
1298 and a specific host). It was also confused by IPv6 addresses, because of
1299 the additional colons they contain. I have fixed the IPv4 problem, and
1300 patched it up to do a reasonable job for IPv6.
1302 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1303 delivery, the log line now contains "pipelined" if PIPELINING was used.
1305 4. An SMTP transport process used to panic and die if the bind() call to set
1306 an explicit outgoing interface failed. This has been changed; it is now
1307 treated in the same way as a connect() failure.
1309 5. A reference to $sender_host_name in the part of a conditional expansion
1310 that was being skipped was still causing a DNS lookup. This no longer
1313 6. The def: expansion condition was not recognizing references to header lines
1314 that used bh_ and bheader_.
1316 7. Added the _cache feature to named lists.
1318 8. The code for checking quota_filecount in the appendfile transport was
1319 allowing one more file than it should have been.
1321 9. For compatibility with Sendmail, the command line option
1329 and sets the incoming protocol and host name (for trusted callers). The
1330 host name and its colon can be omitted when only the protocol is to be set.
1331 Note the Exim already has two private options, -pd and -ps, that refer to
1332 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1333 "s", but I don't think that's a major issue.
1335 10. A number of refactoring changes to the code, none of which should affect
1338 (a) The number of logging options was getting close to filling up the
1339 32-bit word that was used as a bit map. I have split them into two classes:
1340 those that are passed in the argument to log_write(), and those that are
1341 only ever tested independently outside of that function. These are now in
1342 separate 32-bit words, so there is plenty of room for expansion again.
1343 There is no change in the user interface or the logging behaviour.
1345 (b) When building, for example, log lines, the code previously used a
1346 macro that called string_cat() twice, in order to add two strings. This is
1347 not really sufficiently general. Furthermore, there was one instance where
1348 it was actually wrong because one of the argument was used twice, and in
1349 one call a function was used. (As it happened, calling the function twice
1350 did not affect the overall behaviour.) The macro has been replaced by a
1351 function that can join an arbitrary number of extra strings onto a growing
1354 (c) The code for expansion conditions now uses a table and a binary chop
1355 instead of a serial search (which was left over from when there were very
1356 few conditions). Also, it now recognizes conditions like "pam" even when
1357 the relevant support is not compiled in: a suitably worded error message is
1358 given if an attempt is made to use such a condition.
1360 11. Added ${time_interval:xxxxx}.
1362 12. A bug was causing one of the ddress fields not to be passed back correctly
1363 from remote delivery subprocesses. The field in question was not being
1364 subsequently used, so this caused to problems in practice.
1366 13. Added new log selectors queue_time and deliver_time.
1368 14. Might have fixed a bug in maildirsizefile handling that threw up
1369 "unexpected character" debug warnings, and recalculated the data
1370 unnecessarily. In any case, I expanded the warning message to give more
1373 15. Added the message "Restricted characters in address" to the statements in
1374 the default ACL that block characters like @ and % in local parts.
1376 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1377 Three changes have been made:
1379 (a) There was a serious bug; a negative response to MAIL caused the whole
1380 recipient domain to be cached as invalid, thereby blocking all messages
1381 to all local parts at the same domain, from all senders. This bug has
1382 been fixed. The domain is no longer cached after a negative response to
1383 MAIL if the sender used is not empty.
1385 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1388 (c) A new callout option, "use_sender" has been added for people who want
1389 the modified behaviour.
1395 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1398 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1399 indeed breaks things for older releases.
1401 3. Added additional logging to the case where there is a problem reading data
1402 from a filter that is running in a subprocess using a pipe, in order to
1403 try to track down a specific problem.
1405 4. Testing facility fudge: when running in the test harness and attempting
1406 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1407 getting "No route to host". Convert this to a timeout.
1409 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1412 6. Some OS don't have socklen_t but use size_t instead. This affects the
1413 fifth argument of getsockopt() amongst other things. This is now
1414 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1415 can be set for individual OS. I have set it for SunOS5, OSF1, and
1416 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1417 some earlier ones do not.
1419 7. Change 4.30/15 was not doing the test caselessly.
1421 8. The standard form for an IPv6 address literal was being rejected by address
1422 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1423 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1424 this, as well as the form without the "IPv6" on the front (but only when
1425 address literals are enabled, of course).
1427 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1429 10. Exim crashed if a message with an empty sender address specified by -f
1430 encountered a router with an errors_to setting. This could be provoked only
1431 by a command such as
1435 where an empty string was supplied; "<>" did not hit this bug.
1437 11. Installed PCRE release 4.5.
1439 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1440 remained set. It is now erased.
1442 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1443 times from message ids (which are base 36 rather than the normal 62).
1445 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1446 were being counted as actual protocol errors, and logged if the log
1447 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1448 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1449 and DATA following a set of rejected RCPTs do not count as protocol errors.
1450 In other words, Exim assumes they were pipelined, though this may not
1451 actually be the case. Of course, in all cases the client gets an
1452 appropriate error code.
1454 15. If a lookup fails in an ACL condition, a message about the failure may
1455 be available; it is used if testing the ACL cannot continue, because most
1456 such messages specify what the cause of the deferral is. However, some
1457 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1458 that caused an old message to be retained and used if a later statement
1459 caused a defer, replacing the real cause of the deferral.
1461 16. If an IP address had so many PTR records that the DNS lookup buffer
1462 was not large enough to hold them, Exim could crash while trying to process
1463 the truncated data. It now detects and logs this case.
1465 17. Further to 4.21/58, another change has been made: if (and only if) the
1466 first line of a message (the first header line) ends with CRLF, a bare LF
1467 in a subsequent header line has a space inserted after it, so as not to
1468 terminate the header.
1470 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1471 unnecessarily, used atoi() instead of strtol(), and didn't check the
1472 termination when getting file sizes from file names by regex.
1474 19. Completely re-implemented the support for maildirsize files, in the light
1475 of a number of problems with the previous contributed implementation
1476 (4.30/29). In particular:
1478 . If the quota is zero, the maildirsize file is maintained, but no quota is
1481 . If the maildir directory does not exist, it is created before any attempt
1482 to write a maildirsize file.
1484 . The quota value in the file is just a cache; if the quota is changed in
1485 the transport, the new value overrides.
1487 . A regular expression is available for excluding directories from the
1490 20. The autoreply transport checks the characters in options that define the
1491 message's headers; it allows continued headers, but it was checking with
1492 isspace() after an embedded newline instead of explicitly looking for a
1495 21. If all the "regular" hosts to which an address was routed had passed their
1496 expiry times, and had not reached their retry times, the address was
1497 bounced, even if fallback hosts were defined. Now Exim should go on to try
1500 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1501 equivalent code in the SMTP transport. Some hosts send humungous responses
1502 to HELO/EHLO, more than 1024 it seems.
1504 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1505 gives compiler warnings in some environments. I've now done it "properly",
1508 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1509 (because of problems with the built-in one) was declared to return uschar *
1510 instead of char *, causing compiler failure.
1512 25. Fixed a file descriptor leak when processing alias/forward files.
1514 26. Fixed a minor format string issue in dbfn.c.
1516 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1518 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1519 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1522 29. When a sender address is verified, it is cached, to save repeating the test
1523 when there is more than one recipient in a message. However, when the
1524 verification involves a callout, it is possible for different callout
1525 options to be set for different recipients. It is too complicated to keep
1526 track of this in the cache, so now Exim always runs a verification when a
1527 callout is required, relying on the callout cache for the optimization.
1528 The overhead is duplication of the address routing, but this should not be
1531 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1532 to be verified with callout=postmaster, and the main callout worked but the
1533 postmaster check failed, the verification correctly failed. However, if a
1534 subsequent RCPT command asked for sender verification *without* the
1535 postmaster check, incorrect caching caused this verification also to fail,
1538 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1539 it was not caching the DNS options (qualify_single, search_parents) that
1540 were used when the lookup failed. A subsequent lookup with different
1541 options therefore always gave the same answer, though there were cases
1542 where it should not have. (Example: a "domains = !$mx_any" option on a
1543 dnslookup router: the "domains" option is always processed without any
1544 widening, but the router might have qualify_single set.) Now Exim uses the
1545 cached value only when the same options are set.
1547 32. Added John Jetmore's "exipick" utility to the distribution.
1549 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1550 than a timeout (e.g. a certificate is required, and is not provided), an
1551 Exim server now closes the connection immediately. Previously it waited for
1552 the client to close - but if the client is SSL, it seems that they each
1553 wait for each other, leading to a delay before one of them times out.
1555 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1556 maintained 0.8.x compatibility because I don't think many are using it, and
1557 it is clearly obsolete.
1559 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1562 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1563 client certificate was expired. A simple patch fixes this, though I don't
1564 understand the full logic of why the verify callback is called multiple
1567 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1568 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1569 which causes problems with some clients (such as the Certicom SSL Plus
1570 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1571 disables the coutermeasure allowing Eudora to connect."
1573 38. Exim was not checking that a write() to a log file succeeded. This could
1574 lead to Bad Things if a log got too big, in particular if it hit a file
1575 size limit. Exim now panics and dies if it cannot write to a log file, just
1576 as it does if it cannot open a log file.
1578 39. Modified OS/Makefile-Linux so that it now contains
1580 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1582 The two -D definitions ensure that Exim is compiled with large file
1583 support, which makes it possible to handle log files that are bigger than
1586 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1587 instance) a domain was checked against a named list that involved a lookup,
1588 causing $domain_data to be set, then another domain was checked against the
1589 same list, then the first domain was re-checked, the value of $domain_data
1590 after the final check could be wrong. In particular, if the second check
1591 failed, it could be set empty. This bug probably also applied to
1594 41. The strip_trailing_dot option was not being applied to the address given
1595 with the -f command-line option.
1597 42. The code for reading a message's header from the spool was incrementing
1598 $received_count, but never initializing it. This meant that the value was
1599 incorrect (doubled) while delivering a message in the same process in which
1600 it was received. In the most common configuration of Exim, this never
1601 happens - a fresh exec is done - but it can happen when
1602 deliver_drop_privilege is set.
1604 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1605 it now includes up to 150 characters of the unexpected data in the log
1608 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1609 and building data strings. The size of both of these buffers was 10 000
1610 bytes - far larger than anybody would *ever* want, thought I. Needless to
1611 say, somebody hit the limit. I have increased the maximum line length to
1612 20 000 and the maximum data length of concatenated lines to 100 000. I have
1613 also fixed two bugs, because there was no checking on these buffers. Tsk,
1614 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1615 buffer is too small.
1617 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1618 lsearch lookups. Now it does.
1620 46. When parsing a route_list item in a manualroute router, a fixed-length
1621 buffer was used for the list of hosts. I made this 1024 bytes long,
1622 thinking that nobody would ever have a list of hosts that long. Wrong.
1623 Somebody had a whole pile of complicated expansion conditions, and the
1624 string was silently truncated, leading to an expansion error. It turns out
1625 that it is easier to change to an unlimited length (owing to other changes
1626 that have happened since this code was originally written) than to build
1627 structure for giving a limitation error. The length of the item that
1628 expands into the list of hosts is now unlimited.
1630 47. The lsearch lookup could not handle data where the length of text line was
1631 more than 4095 characters. Such lines were truncated, leading to shortened
1632 data being returned. It should now handle lines of any length.
1634 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1635 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1638 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1639 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1640 no longer happens in comment lines. A list of replacements is now placed
1641 at the head of all of the source files, except those whose only change is
1642 to replace PERL_COMMAND in the very first #! line.
1644 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1645 messages on the queue, with a bottom-up merge sort, using code contributed
1646 by Michael Haardt. This should make operations like -bp somewhat faster on
1647 large queues. It won't affect queue runners, except when queue_run_in_order
1650 51. Installed eximstats 1.31 in the distribution.
1652 52. Added support for SRV lookups to the dnslookup router.
1654 53. If an ACL referred to $message_body or $message_body_end, the value was not
1655 reset for any messages that followed in the same SMTP session.
1657 54. The store-handling optimization for building very long strings was not
1658 differentiating between the different store pools. I don't think this
1659 actually made any difference in practice, but I've tidied it.
1661 55. While running the routers to verify a sender address, $sender_address
1662 was still set to the sender address. This is wrong, because when routing to
1663 send a bounce to the sender, it would be empty. Therefore, I have changed
1664 it so that, while verifying a sender address, $sender_address is set to <>.
1665 (There is no change to what happens when verifying a recipient address.)
1667 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1668 target A or AAAA records (if not already returned) without resetting the
1669 qualify_single or search_parents options of the DNS resolver. These are
1670 inappropriate in this case because the targets of MX and SRV records must
1671 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1672 target that, when qualified, matched something in the local domain. These
1673 two options are now turned off when doing these lookups.
1675 57. It seems that at least some releases of Reiserfs (which does not have the
1676 concept of a fixed number of inodes) returns zero and not -1 for the
1677 number of available inodes. This interacted badly with check_spool_inodes,
1678 which assumed that -1 was the "no such thing" setting. What I have done is
1679 to check that the total number of inodes is greater than zero before doing
1680 the test of how many are available.
1682 58. When a "warn" ACL statement has a log_message modifier, the message is
1683 remembered, and not repeated. This is to avoid a lot of repetition when a
1684 message has many recipients that cause the same warning to be written.
1685 Howewer, Exim was preserving the list of already written lines for an
1686 entire SMTP session, which doesn't seem right. The memory is now reset if a
1687 new message is started.
1689 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1690 debugging output unless log_rewrite was also set.
1692 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1693 of (int)(handle) when we know that handle contains (void *)(-1).
1695 61. The Exim daemon panic-logs an error return when it closes the incoming
1696 connection. However "connection reset by peer" seems to be common, and
1697 isn't really an error worthy of noting specially, so that particular error
1700 62. When Exim is trying to find all the local interfaces, it used to panic and
1701 die if the ioctl to get the interface flags failed. However, it seems that
1702 on at least one OS (Solaris 9) it is possible to have an interface that is
1703 included in the list of interfaces, but for which you get a failure error
1704 for this call. This happens when the interface is not "plumbed" into a
1705 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1706 failure of the "get flags" call assumes that the interface is down.
1708 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1709 makes life easier for people who are doing arithmetic on fields extracted
1710 from dates, where you often get leading zeros that should not be
1711 interpreted as octal.
1713 64. Added qualify_domain to the redirect router, to override the global
1716 65. If a pathologically long header line contained very many addresses (the
1717 report of this problem mentioned 10 000) and each of them was rewritten,
1718 Exim could use up a very large amount of memory. (It kept on making new
1719 copies of the header line as it rewrote, and never released the old ones.)
1720 At the expense of a bit more processing, the header rewriting function has
1721 been changed so that it no longer eats memory in this way.
1723 66. The generation of the Received: header has been moved from the time that a
1724 message starts to be received, to the time that it finishes. The timestamp
1725 in the Received: header should now be very close to that of the <= log
1726 line. There are two side-effects of this change:
1728 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1729 logged header lines no longer include the local Received: line, because
1730 it has not yet been created. The same applies to a copy of the message
1731 that is returned to a non-SMTP sender when a message is rejected.
1733 (b) When a filter file is tested using -bf, no additional Received: header
1734 is added to the test message. After some thought, I decided that this
1737 This change does not affect the value of $received_for. It is still set
1738 after address rewriting, but before local_scan() is called.
1740 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1742 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1743 gave an unhelpful panic error message, and a defer error. I have managed to
1744 change this behaviour so that it now rejects any supplied certificate,
1745 which seems right, as the list of acceptable certificates is empty.
1747 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1748 gave an unhelpful defer error. I have not managed to make this reject any
1749 supplied certificates, but the error message it gives is "no certificate
1750 supplied", which is not helpful.
1752 70. exigrep's output now also includes lines that are not associated with any
1753 message, but which match the given pattern. Implemented by a patch from
1754 Martin Sluka, which also tidied up the Perl a bit.
1756 71. Recipient callout verification, like sender verification, was using <> in
1757 the MAIL FROM command. This isn't really the right thing, since the actual
1758 sender may affect whether the remote host accepts the recipient or not. I
1759 have changed it to use the actual sender in the callout; this means that
1760 the cache record is now keyed on a recipient/sender pair, not just the
1761 recipient address. There doesn't seem to be a real danger of callout loops,
1762 since a callout by the remote host to check the sender would use <>.
1763 [SEE ABOVE: changed after hitting problems.]
1765 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1766 temporary errors. However, in the case of such a code being given after
1767 the end of a data transmission (i.e. after ".") Exim was failing to write
1768 a retry record for the message. (Yes, there was some broken host that was
1769 actually sending 8xx at this point.)
1771 73. An unknown lookup type in a host list could cause Exim to panic-die when
1772 the list was checked. (An example that provoked this was putting <; in the
1773 middle of a list instead of at the start.) If this happened during a DATA
1774 ACL check, a -D file could be left lying around. This kind of configuration
1775 error no longer causes Exim to die; instead it causes a defer errror. The
1776 incident is still logged to the main and panic logs.
1778 74. Buglet left over from Exim 3 conversion. The message "too many messages
1779 in one connection" was written to the rejectlog but not the mainlog, except
1780 when address rewriting (yes!) was being logged.
1782 75. Added write_rejectlog option.
1784 76. When a system filter was run not as root (that is, when system_filter_user
1785 was set), the values of the $n variables were not being returned to the
1786 main process; thus, they were not subsequently available in the $sn
1789 77. Added +return_path_on_delivery log selector.
1791 78. A connection timeout was being treated differently from recipients deferred
1792 when testing hosts_max_try with a message that was older than the host's
1793 retry timeout. (The host should not be counted, thus allowing all hosts to
1794 be tried at least once before bouncing.) This may have been the cause of an
1795 occasionally reported bug whereby a message would remain on the queue
1796 longer than the retry timeout, but would be bounced if a delivery was
1797 forced. I say "may" because I never totally pinned down the problem;
1798 setting up timeout/retry tests is difficult. See also the next item.
1800 79. The ultimate address timeout was not being applied to errors that involved
1801 a combination of host plus message (for example, a timeout on a MAIL
1802 command). When an address resolved to a number of possible hosts, and they
1803 were not all tried for each delivery (e.g. because of hosts_max_try), a
1804 message could remain on the queue longer than the retry timeout.
1806 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1809 81. Fixed an obscure SMTP outgoing bug which required at least the following
1810 conditions: (a) there was another message waiting for the same server;
1811 (b) the server returned 5xx to all RCPT commands in the first message so
1812 that the message was not completed; (c) the server dropped the connection
1813 or gave a negative response to the RSET that Exim sends to abort the
1814 transaction. The observed case was a dropped connection after DATA that had
1815 been sent in pipelining mode. That is, the server had advertised PIPELINING
1816 but was not implementing it correctly. The effect of the bug was incorrect
1817 behaviour, such as trying another host, and this could lead to a crash.
1823 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1824 and daemon.c were passed as pointers to ints; they should have been
1825 pointers to socklen_t variables (which are typically unsigned ints).
1827 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1830 3. Fixed a really odd bug that affected only the testing scheme; patching a
1831 certain fixed string in the binary changed the value of another string that
1832 happened to be identical to the end of the original first string.
1834 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1835 name", it returns that address as the IP address. On some operating
1836 systems (e.g. Solaris), it also passes back the IP address string as the
1837 "host name". However, on others (e.g. Linux), it passes back an empty
1838 string. Exim wasn't checking for this, and was changing the host name to an
1839 empty string, assuming it had been canonicized.
1841 5. Although rare, it is permitted to have more than one PTR record for a given
1842 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1843 all the names associated with an address, because they do in Solaris.
1844 However, it seems that they do not in Linux for data that comes from the
1845 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1846 I found this out when I moved to a new Linux workstation and tried to run
1847 the Exim test suite.
1849 To get round this problem I have changed the code so that it now does its
1850 own call to the DNS to look up PTR records when searching for a host name.
1851 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1852 addresses that are only in /etc/hosts are still found.
1854 This behaviour is, however, controlled by an option called host_lookup_
1855 order, which defaults to "bydns:byaddr". If people want to use the other
1856 order, or indeed, just use one or the other means of lookup, they can
1857 specify it in this variable.
1859 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1860 some operating systems, this comes back from gethostbyaddr() as an empty
1861 string, and this is what Exim used to test for. However, it seems that in
1862 other systems, "." is yielded. Exim now tests for this case too.
1864 7. The values of check_spool_space and check_log_space are now held internally
1865 as a number of kilobytes instead of an absolute number of bytes. If a
1866 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1867 kilobyte. This means that much larger values can be stored.
1869 8. Exim monitor: an attempt to get the action menu when not actually pointing
1870 at a message produces an empty menu entitled "No message selected". This
1871 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1872 no entries in it ("Shell widget menu has zero width and/or height"). So I
1873 have added a single, blank menu entry in this case.
1875 9. Added ${quote_local_part.
1877 10. MIME decoding is now applied to the contents of Subject: header lines when
1880 11. Now that a reference to $sender_host_address automatically causes a reverse
1881 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1882 host lookup before query-style lookups in lists that might use this
1883 variable. This has therefore been abolished, and the "net-" prefix is no
1884 longer necessary for query-style lookups.
1886 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1887 have been a typo for LFLAGS, so it has been changed.
1889 13. The install script calls Exim with "-C /dev/null" in order to find the
1890 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1891 to be output. Howeve, since Exim outputs its version number before the
1892 error, it didn't break the script. It just looked ugly. I fixed this by
1893 always allowing "-C /dev/null" if the caller is root.
1895 14. Ignore overlarge ACL variable number when reading spool file - insurance
1896 against a later release with more variables having written the file.
1898 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1899 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1900 this, as well as the form without the "IPv6" on the front.
1902 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1903 OS/Makefile-Darwin file.
1905 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1906 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1908 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1909 variable. However, some OS use ints for inodes. I've added cast to long int
1910 to get rid of the compiler warning.
1912 19. I had forgotten to lock out "/../" in configuration file names when
1913 ALT_CONFIG_PREFIX was set.
1915 20. Routers used for verification do not need to specify transports. However,
1916 if such a router generated a host list, and callout was configured, Exim
1917 crashed, because it could not find a port number from the (non-existent)
1918 transport. It now assumes port 25 in this circumstance.
1920 21. Added the -t option to exigrep.
1922 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1923 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1924 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1925 them is set, LOOKUP_LSEARCH is forced.
1927 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1929 24. Added sender and host information to the "rejected by local_scan()" log
1930 line; previously there was no indication of these.
1932 25. Added .include_if_exists.
1934 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1935 Linux. It turns out that not all file systems support this. Apparently some
1936 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1937 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1938 not supported on the file descriptor, is now ignored when Exim is trying to
1939 sync a directory. This applies only to Linux.
1941 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1943 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1944 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1947 29. Added support for maildirsize files from supplied patch (modified a bit).
1949 30. The use of :fail: followed by an empty string could lead Exim to respond to
1950 sender verification failures with (e.g.):
1952 550 Verification failed for <xxx>
1953 550 Sender verify failed
1955 where the first response line was missing the '-' that indicates it is not
1956 the final line of the response.
1958 31. The loop for finding the name of the user that called Exim had a hardwired
1959 limit of 10; it now uses the value of finduser_retries, which is used for
1960 all other user lookups.
1962 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1965 33. Exim was neglecting to zero errno before one call of strtol() when
1966 expanding a string and expecting an integer value. On some systems this
1967 resulted in spurious "integer overflow" errors. Also, it was casting the
1968 result into an int without checking.
1970 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1971 did not work. The code looks as if it has *never* worked, though it appears
1972 to have been documented since at least releast 1.62. I have made it work.
1974 35. The "timeout_DNS" error in retry rules, also documented since at least
1975 1.62, also never worked. As it isn't clear exactly what this means, and
1976 clearly it isn't a major issue, I have abolished the feature by treating it
1977 as "timeout", and writing a warning to the main and panic logs.
1979 36. The display of retry rules for -brt wasn't always showing the error code
1982 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1983 timeout_connect_A, timeout_connect_MX.
1985 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1986 to the empty sender.
1988 39. The daemon was not analysing the content of -oX till after it had closed
1989 stderr and disconnected from the controlling terminal. This meant that any
1990 syntax errors were only noted on the panic log, and the return code from
1991 the command was 0. By re-arranging the code a little, I've made the
1992 decoding happen first, so such errors now appear on stderr, and the return
1993 code is 1. However, the actual setting up of the sockets still happens in
1994 the disconnected process, so errors there are still only recorded on the
1997 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1998 connections (as happens on some IP stacks) was logged at start up time as
1999 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2000 it from "IPv6 and IPv4", which means that two separate sockets are being
2003 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2004 says whether AF_INET or AF_INET6 was passed as an argument.
2006 42. Exiwhat output was messed up when time zones were included in log
2009 43. Exiwhat now gives more information about the daemon's listening ports,
2010 and whether -tls-on-connect was used.
2012 44. The "port" option of the smtp transport is now expanded.
2014 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2015 silently ignored. Now an error message is written to the main and panic
2018 46. There's a new ACL modifier called "logwrite" which writes to a log file
2019 as soon as it is encountered.
2021 47. Added $local_user_uid and $local_user_gid at routing time.
2023 48. Exim crashed when trying to verify a sender address that was being
2026 49. Exim was recognizing only a space character after ".include". It now also
2027 recognizes a tab character.
2029 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2030 extracting the relevant information from the specification. The man page no
2031 longer contains scrambled data for the -d option, and I've added a section
2032 at the front about calling Exim under different names.
2034 51. Added "extra_headers" argument to the "mail" command in filter files.
2036 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2039 53. Installed eximstats 1.29.
2041 54. Added transport_filter_timeout as a generic transport option.
2043 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2044 Cc: header lines. This was required by RFC 822, but it not required by RFC
2047 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2048 incoming messages that did not have them. Now it does so only if the
2049 message originates locally, that is, if there is no associated remote host
2050 address. When Resent- header lines are present, this applies to the Resent-
2051 lines rather than the non-Resent- lines.
2053 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2054 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2056 58. Messages for configuration errors now include the name of the main
2057 configuration file - useful now that there may be more than one file in a
2058 list (.included file names were always shown).
2060 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2061 for those rare installations that do not start the daemon as root or run it
2062 setuid root. I've cut out the call to initgroups() if the daemon is not
2065 60. The Exim user and group can now be bound into the binary as text strings
2066 that are looked up at the start of Exim's processing.
2068 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2070 62. Added $mailstore_basename variable.
2072 63. Installed patch to sieve.c from Michael Haardt.
2074 64. When Exim failed to open the panic log after failing to open the main log,
2075 the original message it was trying to log was written to stderr and debug
2076 output, but if they were not available (the usual case in production), it
2077 was lost. Now it is written to syslog before the two lines that record the
2078 failures to open the logs.
2080 65. Users' Exim filters run in subprocesses under the user's uid. It is
2081 possible for a "deliver" command or an alias in a "personal" command to
2082 provoke an address rewrite. If logging of address rewriting is configured,
2083 this fails because the process is not running as root or exim. There may be
2084 a better way of dealing with this, but for the moment (because 4.30 needs
2085 to be released), I have disabled address rewrite logging when running a
2086 filter in a non-root, non-exim process.
2092 1. The buildconfig auxiliary program wasn't quoting the value set for
2093 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2094 not defined. This bug was masked in 4.22 by the effect that was fixed in
2097 2. Some messages that were rejected after a message id was allocated were
2098 shown as "incomplete" by exigrep. It no longer does this for messages that
2099 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2101 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2102 have allow_domain_literals set, the ID did not get logged in the <= line.
2103 Domain literals are now always recognized in Message-ID: header lines.
2105 4. The first argument for a ${extract expansion item is the key name or field
2106 number. Leading and trailing spaces in this item were not being ignored,
2107 causing some misleading effects.
2109 5. When deliver_drop_privilege was set, single queue runner processes started
2110 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2111 same command in the process it spins off) were not dropping privilege.
2113 6. When the daemon running as "exim" started a queue runner, it always
2114 re-executed Exim in the spun-off process. This is a waste of effort when
2115 deliver_drop_privilege is set. The new process now just calls the
2116 queue-runner function directly.
2122 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2125 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2126 ignored. Though the use of -oP was forcing the writing of a pid file, it
2127 was always written to the default place.
2129 3. If the message "no IP address found for host xxxx" is generated during
2130 incoming verification, it is now followed by identification of the incoming
2131 connection (so you can more easily find what provoked it).
2133 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2135 5. Added some features to "harden" Exim a bit more against certain attacks:
2137 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2138 be put in Local/Makefile. This is like the never_users runtime option,
2139 but it cannot be overridden. The default setting is "root".
2141 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2142 prefix string with which any file named in a -C command line option
2145 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2146 is retained for -C and -D only if the caller of Exim is root. Without
2147 it, the exim user may also use -C and -D and retain privilege.
2149 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2150 command line option is disabled.
2152 6. Macro names set by the -D option must start with an upper case letter, just
2153 like macro names defined in the configuration file.
2155 7. Added "dereference=" facility to LDAP.
2157 8. Two instances of the typo "uknown" in the source files are fixed.
2159 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2160 the Configure-Makefile script screwed up while processing it.
2162 10. Incorporated PCRE 4.4.
2164 11. The SMTP synchronization check was not operating right at the start of an
2165 SMTP session. For example, it could not catch a HELO sent before the client
2166 waited for the greeting. There is now a check for outstanding input at the
2167 point when the greeting is written. Because of the duplex, asynchronous
2168 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2169 way, but not yet received, when the check is performed.
2171 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2172 on TCP/IP sockets, because this apparently causes some broken clients to
2175 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2176 unchanged) from the Cygwin maintainer.
2178 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2179 was supported instead of testing for mbx. Effectively a typo.
2181 15. The spa authenticator server code was not checking that the input it
2182 received was valid base64.
2184 16. The debug output line for the "set" modifier in ACLs was not showing the
2185 name of the variable that was being set.
2187 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2189 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2190 lookup on its own. Something else was needed to trigger the lookup. For
2191 example, a match in host_lookup or the need for a host name in a host list.
2192 Now, if $sender_host_name is referenced and the host name has not yet been
2193 looked up, a lookup is performed. If the lookup fails, the variable remains
2194 empty, and $host_lookup_failed is set to "1".
2196 19. Added "eqi" as a case-independent comparison operator.
2198 20. The saslauthd authentication condition could segfault if neither service
2199 nor realm was specified.
2201 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2202 error message that was logged was misleading, and incoming SMTP
2203 connections were dropped. The message is now more accurate, and temporary
2204 errors are given to SMTP connections.
2206 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2207 (except RSET) with a 421 error, until QUIT is received. However, it was
2208 failing to send a response to QUIT.
2210 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2211 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2212 ACL is now run after the helo_try_verify_hosts code.
2214 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2215 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2216 case-independent, so other case variants are also recognized). Apparently
2217 some systems use these upper case variants.
2219 25. If more than two messages were waiting for the same host, and a transport
2220 filter was specified for the transport, Exim sent two messages over the
2221 same TCP/IP connection, and then failed with "socket operation on non-
2222 socket" when it tried to send the third.
2224 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2226 27. The extern definition of crypt16() in expand.c was not being excluded when
2227 the OS had its own crypt16() function.
2229 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2230 as a preferred synonym for return_size_limit, both as an option and as an
2233 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2235 30. Changed the default configuration ACL to relax the local part checking rule
2236 for addresses that are not in any local domains. For these addresses,
2237 slashes and pipe symbols are allowed within local parts, but the sequence
2238 /../ is explicitly forbidden.
2240 31. SPA server authentication was not clearing the challenge buffer before
2243 32. log_message in a "warn" ACL statement was writing to the reject log as
2244 well as to the main log, which contradicts the documentation and doesn't
2245 seem right (because no rejection is happening). So I have stopped it.
2247 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2248 However, I am unable to do any testing of this.
2250 34. Fixed an infelicity in the appendfile transport. When checking directories
2251 for a mailbox, to see if any needed to be created, it was accidentally
2252 using path names with one or more superfluous leading slashes; tracing
2253 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2255 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2256 contents are added to the log line that is written for every discarded
2257 recipient. (Previously a log_message setting was ignored.)
2259 36. The ${quote: operator now quotes the string if it is empty.
2261 37. The install script runs exim in order to find its version number. If for
2262 some reason other than non-existence or emptiness, which it checks, it
2263 could not run './exim', it was installing it with an empty version number,
2264 i.e. as "exim-". This error state is now caught, and the installation is
2267 38. An argument was missing from the function that creates an error message
2268 when Exim fails to connect to the socket for saslauthd authentication.
2269 This could cause Exim to crash, or give a corrupted message.
2271 39. Added isip, isip4, and isip6 to ${if conditions.
2273 40. The ACL variables $acl_xx are now saved with the message, and can be
2274 accessed later in routers, transports, and filters.
2276 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2277 strings in the file are not string-expanded.
2279 42. If a MAIL command specified a SIZE value that was too large to fit into an
2280 int variable, the check against message_size_limit failed. Such values are
2281 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2282 day this will have to be increased, but I don't think I want to be around
2283 when emails are that large.
2290 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2291 iconv() is not standard in FreeBSD.
2293 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2294 IPv6 enabled. The observed symptom was a segmentation fault on return from
2295 the function os_common_find_running_interfaces() in src/os.c.
2297 3. In the check_special_case() function in daemon.c I had used "errno" as an
2298 argument name, which causes warnings on some systems. This was basically a
2299 typo, since it was named "eno" in the comments!
2301 4. The code that waits for the clock to tick (at a resolution of some fraction
2302 of a second) so as to ensure message-id uniqueness was always waiting for
2303 at least one whole tick, when it could have waited for less. [This is
2304 almost certainly not relevant at current processor speeds, where it is
2305 unlikely to ever wait at all. But we try to future-proof.]
2307 5. The function that sleeps for a time interval that includes fractions of a
2308 second contained a race. It did not block SIGALRM between setting the
2309 timer, and suspending (a couple of lines later). If the interval was short
2310 and the sigsuspend() was delayed until after it had expired, the suspension
2311 never ended. On busy systems this could lead to processes getting stuck for
2314 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2315 process, before it forks any delivery processes. The open lookup caching
2316 mechanism meant that the open file or database connection was passed into
2317 the delivery process. The problem was that delivery processes always tidy
2318 up cached lookup data. This could cause a problem for the next delivery
2319 process started by the queue runner, because the external queue runner
2320 process does not know about the closure. So the next delivery process
2321 still has data in the lookup cache. In the case of a file lookup, there was
2322 no problem because closing a file descriptor in a subprocess doesn't affect
2323 the parent. However, if the lookup was caching a connection to a database,
2324 the connection was closed, and the second delivery process was likely to
2325 see errors such as "PGSQL: query failed: server closed the connection
2326 unexpectedly". The problem has been fixed by closing all cached lookups
2327 in a queue runner before running a delivery process.
2329 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2330 seem to have the "const" qualifier which it has on other OS. I've
2333 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2334 *of the same type* (client or server) with the same public name that an
2335 error should be diagnosed.
2337 9. When Exim looked up a host name for an IP address, but failed to find the
2338 original IP address when looking up the host name (a safety check), it
2339 output the message "<ip address> does not match any IP for NULL", which was
2340 confusing, to say the least. The bug was that the host name should have
2341 appeared instead of "NULL".
2343 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2344 user that is built into the binary, and the -C or -D options is used, root
2345 privilege is dropped before the configuration file is read. In addition,
2346 logging is switched to stderr instead of the normal log files. If the
2347 configuration then re-defines the Exim user, the unprivileged environment
2348 is probably not what is expected, so Exim logs a panic warning message (but
2351 However, if deliver_drop_privilege is set, the unprivileged state may well
2352 be exactly what is intended, so the warning has been cut out in that case,
2353 and Exim is allowed to try to write to its normal log files.
2359 1. smtp_return_error_details was not giving details for temporary sender
2360 or receiver verification errors.
2362 2. Diagnose a configuration error if two authenticators have the same public
2365 3. Exim used not to create the message log file for a message until the first
2366 delivery attempt. This could be confusing when incoming messages were held
2367 for policy or load reasons. The message log file is now created at the time
2368 the message is received, and an initial "Received" line is written to it.
2370 4. The automatically generated man page for command line options had a minor
2371 bug that caused no ill effects; however, a more serious problem was that
2372 the procedure for building the man page automatically didn't always
2373 operate. Consequently, release 4.20 contains an out-of-date version. This
2374 shouldn't happen again.
2376 5. When building Exim with embedded Perl support, the script that builds the
2377 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2378 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2380 6. The freeze_tell option was not being used for messages that were frozen on
2381 arrival, either by an ACL or by local_scan().
2383 7. Added the smtp_incomplete_transaction log selector.
2385 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2386 was accepting AUTH without a new EHLO.
2388 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2389 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2392 10. Logging of TCP/IP connections (when configured) now happens in the main
2393 daemon process instead of the child process, so that the TCP/IP connection
2394 count is more accurate (but it can never be perfect).
2396 11. The use of "drop" in a nested ACL was not being handled correctly in the
2397 outer ACL. Now, if condition failure induced by the nested "drop" causes
2398 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2399 or "require"), the connection is dropped.
2401 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2402 that yield "discard" can now be used with an "accept" or a "discard" verb,
2403 but an error is generated for any others (because I can't see a useful way
2404 to define what should happen).
2406 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2407 are now processed in the same way as lines in the Exim configuration file.
2408 In particular, continuation lines are supported.
2410 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2412 15. Added -ti meaning -t -i.
2414 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2415 domains, and warn by logging if others are found.
2417 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2418 returned from GIFCONF when Exim is trying to find the list of interfaces on
2419 a host. The code in os.c has been modified to copy each ifreq to an aligned
2420 structure in all cases.
2422 Also, in some cases, the returned ifreq's were being copied to a 'struct
2423 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2424 means the last couple of bytes of an IPv6 address could be chopped if the
2425 ifreq contained only a normal sockaddr (14 bytes storage).
2427 18. Named domain lists were not supported in the hosts_treat_as_local option.
2428 An entry such as +xxxx was not recognized, and was treated as a literal
2431 19. Ensure that header lines added by a DATA ACL are included in the reject log
2432 if the ACL subsequently rejects the message.
2434 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2435 MD5 (which is deprecated).
2437 21. When testing a filter file using -bf, Exim was writing a message when it
2438 took the sender from a "From " line in the message, but it was not doing so
2439 when it took $return_path from a Return-Path: header line. It now does.
2441 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2442 with a valid header line field name (a series of printing characters
2443 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2445 23. Changed "disc" in the source to "disk" to conform to the documentation and
2446 the book and for uniformity.
2448 24. Ignore Sendmail's -Ooption=value command line item.
2450 25. When execve() failed while trying to run a command in a pipe transport,
2451 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2452 could be confused with a return value of 69 from the command itself. This
2453 has been changed to 127, the value the shell returns if it is asked to run
2454 a non-existent command. The wording for the related log line suggests a
2455 non-existent command as the problem.
2457 26. If received_header_text expands to an empty string, do not add a Received:
2458 header line to the message. (Well, it adds a token one on the spool, but
2459 marks it "old" so that it doesn't get used or transmitted.)
2461 27. Installed eximstats 1.28 (addition of -nt option).
2463 28. There was no check for failure on the call to getsockname() in the daemon
2464 code. This can fail if there is a shortage of resources on the system, with
2465 ENOMEM, for example. A temporary error is now given on failure.
2467 29. Contrary to the C standard, it seems that in some environments, the
2468 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2469 program. Exim now does this explicitly; it affects the formatting of
2470 timestamps using strftime().
2472 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2473 complaints. I've now initialized all the variables, to avoid this.
2475 32. Header lines added by a system filter were not being "seen" during
2476 transport-time rewrites.
2478 33. The info_callback() function passed to OpenSSL is set up with type void
2479 (*)(SSL *, int, int), as described somewhere. However, when calling the
2480 function (actually a macro) that sets it up, the type void(*)() is
2481 expected. I've put in a cast to prevent warnings from picky compilers.
2483 34. If a DNS black list lookup found a CNAME record, but there were no A
2484 records associated with the domain it pointed at, Exim crashed.
2486 35. If a DNS black list lookup returned more than one A record, Exim ignored
2487 all but the first. It now scans all returned addresses if a particular IP
2488 value is being sought. In this situation, the contents of the
2489 $dnslist_value variable are a list of all the addresses, separated by a
2492 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2493 to accept a host name and all its aliases if the forward lookup for any of
2494 them yielded the IP address of the incoming connection. Now it accepts only
2495 those names whose forward lookup yields the correct IP address. Any other
2496 names are discarded. This closes a loophole whereby a rogue DNS
2497 administrator could create reverse DNS records to break through a
2498 wildcarded host restriction in an ACL.
2500 37. If a user filter or a system filter that ran in a subprocess used any of
2501 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2502 the wrong values were passed to the pipe command ($thisaddress had the
2503 value of $0, $0 had the value of $1, etc). This bug was introduced by
2504 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2506 38. Improved the line breaking for long SMTP error messages from ACLs.
2507 Previously, if there was no break point between 40 and 75 characters, Exim
2508 left the rest of the message alone. Two changes have been made: (a) I've
2509 reduced the minimum length to 35 characters; (b) if it can't find a break
2510 point between 35 and 75 characters, it looks ahead and uses the first one
2511 that it finds. This may give the occasional overlong line, but at least the
2512 remaining text gets split now.
2514 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2515 file descriptors might be low, and that setting 1000 would always raise it.
2516 It turns out that in some environments, the limit is already over 1000 and
2517 that lowering it causes trouble. So now Exim takes care not to decrease it.
2519 40. When delivering a message, the value of $return_path is set to $sender_
2520 address at the start of routing (routers may change the value). By an
2521 oversight, this default was not being set up when an address was tested by
2522 -bt or -bv, which affected the outcome if any router or filter referred to
2525 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2526 log only when "message" or "log_message" are set. However, if one of the
2527 conditions was an address verification, or a call to a nested ACL, the
2528 messages generated by the underlying test were being passed through. This
2529 no longer happens. The underlying message is available in $acl_verify_
2530 message for both "message" and "log_message" expansions, so it can be
2531 passed through if needed.
2533 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2534 new expansion $bh_ to give the encoded byte string without charset
2535 translation. Translation happens only if iconv() is available; HAVE_ICONV
2536 indicates this at build time. HEADERS_CHARSET gives the charset to
2537 translate to; headers_charset can change it in the configuration, and
2538 "headers charset" can change it in an individual filter file.
2540 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2541 that creates RFC 2047 encoded "words" labels them as that charset instead
2542 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2543 expansion operator; (ii) when Exim creates a From: line for a local
2544 message; (iii) when a header line is rewritten to include a "phrase" part.
2546 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2547 buggy, causing it to skip the first lines of messages whose message ID
2548 ended in 'D'. This would not have bitten before Exim release 4.14, because
2549 message IDs were unlikely to end in 'D' before then. The effect was to have
2550 incorrect size information for certain domains.
2552 45. #include "config.h" was missing at the start of the crypt16.c module. This
2553 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2556 46. If there was a timeout during a "random" callout check, Exim treated it as
2557 a failure of the random address, and carried on sending RSET and the real
2558 address. If the delay was just some slowness somewhere, the response to the
2559 original RCPT would be taken as a response to RSET and so on, causing
2560 mayhem of various kinds.
2562 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2563 when I implemented it. It didn't allow for the fact that some option values
2564 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2565 the right test for positive values.
2567 48. Domain names in DNS records are case-independent. Exim always looks them up
2568 in lower case. Some resolvers return domain names in exactly the case they
2569 appear in the zone file, that is, they may contain uppercase letters. Not
2570 all resolvers do this - some return always lower case. Exim was treating a
2571 change of case by a resolver as a change of domain, similar to a widening
2572 of a domain abbreviation. This triggered its re-routing code and so it was
2573 trying to route what was effectively the same domain again. This normally
2574 caused routing to fail (because the router wouldn't handle the domain
2575 twice). Now Exim checks for this case specially, and just changes the
2576 casing of the domain that it ultimately uses when it transmits the message
2579 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2582 50. If a filter generated a file delivery with a non-absolute name (possible if
2583 no home directory exists for the router), the forbid_file option was not
2586 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2587 the existing equality matching.
2589 52. Exim was using ints instead of ino_t variables in some places where it was
2590 dealing with inode numbers.
2592 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2593 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2594 called TMPDIR, and if it finds it is different, it changes its value.
2596 54. The smtp_printf() function is now made available to local_scan() so
2597 additional output lines can be written before returning. There is also an
2598 smtp_fflush() function to enable the detection of a dropped connection.
2599 The variables smtp_input and smtp_batched_input are exported to
2602 55. Changed the default runtime configuration: the message "Unknown user"
2603 has been removed from the ACL, and instead placed on the localuser router,
2604 using the cannot_route_message feature. This means that any verification
2605 failures that generate their own messages won't get overridden. Similarly,
2606 the "Unrouteable address" message that was in the ACL for unverifiable
2607 relay addresses has also been removed.
2609 56. Added hosts_avoid_esmtp to the smtp transport.
2611 57. The exicyclog script was not checking for the esoteric option
2612 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2613 will work only if exicyclog is run under the appropriate euid.
2615 58. Following a discussion on the list, the rules by which Exim recognises line
2616 endings on incoming messages have been changed. The -dropcr and drop_cr
2617 options are now no-ops, retained only for backwards compatibility. The
2618 following line terminators are recognized: LF CRLF CR. However, special
2619 processing applies to CR:
2621 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2622 nor a local message in the state where . is a terminator.
2624 (ii) If a bare CR is encountered in a header line, an extra space is added
2625 after the line terminator so as not to end the header. The reasoning
2626 behind this is that bare CRs in header lines are most likely either
2627 to be mistakes, or people trying to play silly games.
2629 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2630 was being incorrectly given as 18 bytes larger than it should have been.
2631 This is a VOB (very old bug).
2633 60. This may never have affected anything current, but just in case it has:
2634 When the local host is found other than at the start of a list of hosts,
2635 the local host, those with the same MX, and any that follow, are discarded.
2636 When the list in question was part of a longer list of hosts, the following
2637 hosts (not currently being processed) were also being discarded. This no
2638 longer happens. I'm not sure if this situation could ever has previously
2641 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2644 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2645 header. This is something that is recommended in a new Internet Draft, and
2646 is something that is documented as being done by Sendmail. There are two
2647 possible values. For messages generated by the autoreply transport, Exim
2650 Auto-Submitted: auto-replied
2652 whereas for all other generated messages (e.g. bounces) it adds
2654 Auto-Submitted: auto-generated
2656 63. The "personal" condition in filters now includes a test for the
2657 Auto-Submitted: header. If it contains the string "auto-" the message it
2658 not considered personal.
2660 64. Added rcpt_include_affixes as a generic transport option.
2662 65. Added queue_only_override (default true).
2664 66. Added the syslog_duplication option.
2666 67. If what should have been the first header line of a message consisted of
2667 a space followed by a colon, Exim was mis-interpreting it as a header line.
2668 It isn't of course - it is syntactically invalid and should therefore be
2669 treated as the start of the message body. The misbehaviour could have
2670 caused a number of strange effects, including loss of data in subsequent
2671 header lines, and spool format errors.
2673 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2674 client host had authenticated. This control can now be exercised by an ACL
2675 for more flexibility.
2677 69. By default, callouts do not happen when testing with -bh. There is now a
2678 variant, -bhc, which does actually run the callout code, including
2679 consulting and updating the callout cache.
2681 70. Added support for saslauthd authentication, courtesy of Alexander
2684 71. If statvfs() failed on the spool or log directories while checking their
2685 size for availability, Exim confusingly gave the error "space shortage".
2686 Furthermore, in debugging mode it crashed with a floating point exception.
2687 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2688 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2689 serious problem, Exim now writes to the main and panic logs when this
2690 happens, with details of the failure. It then refuses to accept the
2691 incoming message, giving the message "spool directory problem" or "log
2692 directory problem" with a 421 code for SMTP messages.
2694 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2695 0, 1, and 2 exist, because some OS complain for execs without them (see
2696 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2697 descriptors. However, the code omitted to check that the open succeeded,
2698 causing mysterious errors if for some reason the permissions on /dev/null
2699 got screwed. Now Exim writes a message to the main and panic logs, and
2700 bombs out if it can't open /dev/null.
2702 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2703 interact so that it is all more flexible. It is supposed to remain
2704 backwards compatible. Also added extra_local_interfaces.
2706 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2707 to bomb out with an assertion failure - to the client this appears as a
2708 connection drop. This problem occurs in the part of the code that was taken
2709 from the Samba project. Fortunately, the assertion is in a very simple
2710 function, so I have fixed this by reproducing the function inline in the
2711 one place where it is called, and arranging for authentication to fail
2712 instead of killing the process with assert().
2714 75. The SPA client code was not working when the server requested OEM rather
2715 than Unicode encoding.
2717 76. Added code to make require_files with a specific uid setting more usable in
2718 the case where statting the file as root fails - usually a non-root-mounted
2719 NFS file system. When this happens and the failure is EACCES, Exim now
2720 forks a subprocess and does the per-uid checking as the relevant uid.
2722 77. Added process_log_path.
2724 78. If log_file_path was not explicitly set, a setting of check_log_space or
2725 check_log_inodes was ignored.
2727 79. If a space check for the spool or log partitions fails, the incident is now
2728 logged. Of course, in the latter case the data may get lost...
2730 80. Added the %p formatting code to string_format() so that it can be used to
2731 print addresses in debug_print(). Adjusted all the address printing in the
2732 debugging in store.c to use %p rather than %d.
2734 81. There was a concern that a line of code in smtp_in.c could overflow a
2735 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2736 initially expressed, the concern was not well-founded, because trailing
2737 spaces are removed early. However, if the trailing spaces were followed by
2738 a NULL, they did not get removed, so the overflow was possible. Two fixes
2741 (a) I re-wrote the offending code in a cleaner fashion.
2742 (b) If an incoming SMTP command contains a NULL character, it is rejected
2745 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2746 runs initgroups(), so that if the Exim user is in any additional groups,
2747 they will be used during message reception.
2753 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob