1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.113 2005/04/06 10:53:47 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
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.
128 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
129 hadn't been configured. The fix is from Juergen Kreileder, who
130 understands it better than I do:
132 "Here's what I see happening with three configured cyrus_sasl
133 authenticators configured (plain, login, cram-md5):
135 On startup auth_cyrus_sasl_init() gets called for each of these.
136 This means three calls to sasl_listmech() without a specified mech_list.
137 => SASL tests which mechs of all available mechs actually work
138 => three warnings about OTP not working
139 => the returned list contains: plain, login, cram-md5, digest-md5, ...
141 With the patch, sasl_listmech() also gets called three times. But now
142 SASL's mech_list option is set to the server_mech specified in the the
143 authenticator. Or in other words, the answer from sasl_listmech()
144 gets limited to just the mech you're testing for (which is different
146 => the return list contains just 'plain' or 'login', 'cram-md5' or
147 nothing depending on the value of ob->server_mech.
149 I've just tested the patch: Authentication still works fine,
150 unavailable mechs specified in the exim configuration are still
151 caught, and the auth.log warnings about OTP are gone."
153 PH/25 When debugging is enabled, the contents of the command line are added
154 to the debugging output, even when log_selector=+arguments is not
157 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
158 answer is "GNU", and only if the return is "GNU/something" is the answer
162 A note about Exim versions 4.44 and 4.50
163 ----------------------------------------
165 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
166 changes of various kinds. As a consequence, a big documentation update was
167 needed. This delayed the release for rather longer than seemed good, especially
168 in the light of a couple of (minor) security issues. Therefore, the changes
169 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
170 release. So 4.44 and 4.50 are in effect two different branches that both start
173 I have left the 4.50 change log unchanged; it contains all the changes since
174 4.43. The change log for 4.44 is below; many of its items are identical to
175 those for 4.50. This seems to be the most sensible way to preserve the
176 historical information.
182 1. Minor wording change to the doc/README.SIEVE file.
184 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
185 computation of the current number of files was incorrect.
187 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
188 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
191 4. Give more explanation in the error message when the command for a transport
192 filter fails to execute.
194 5. There are several places where Exim runs a non-Exim command in a
195 subprocess. The SIGUSR1 signal should be disabled for these processes. This
196 was being done only for the command run by the queryprogram router. It is
197 now done for all such subprocesses. The other cases are: ${run, transport
198 filters, and the commands run by the lmtp and pipe transports.
200 6. Added CONFIGURE_GROUP build-time option.
202 7. Some older OS have a limit of 256 on the maximum number of file
203 descriptors. Exim was using setrlimit() to set 1000 as a large value
204 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
205 systems. I've change it so that if it can't get 1000, it tries for 256.
207 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
208 was an oversight, and furthermore, ever since the addition of extra
209 controls (e.g. 4.43/32), the checks on when to allow different forms of
210 "control" were broken. There should now be diagnostics for all cases when a
211 control that does not make sense is encountered.
213 9. Added the /retain_sender option to "control=submission".
215 10. $recipients is now available in the predata ACL (oversight).
217 11. Tidy the search cache before the fork to do a delivery from a message
218 received from the command line. Otherwise the child will trigger a lookup
219 failure and thereby defer the delivery if it tries to use (for example) a
220 cached ldap connection that the parent has called unbind on.
222 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
223 of $address_data from the recipient verification was clobbered by the
226 13. The value of address_data from a sender verification is now available in
227 $sender_address_data in subsequent conditions in the ACL statement.
229 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
231 15. Added a new option "connect=<time>" to callout options, to set a different
234 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
235 was its contents. (It was OK if the option was not defined at all.)
237 17. A "Completed" log line is now written for messages that are removed from
238 the spool by the -Mrm option.
240 18. New variables $sender_verify_failure and $recipient_verify_failure contain
241 information about exactly what failed.
243 19. Added -dd to debug only the daemon process.
245 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
246 handles timeouts, both on the server side and network timeouts. Renamed the
247 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
249 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
250 It is now set to "smtps".
252 22. $host_address is now set to the target address during the checking of
255 23. When checking ignore_target_hosts for an ipliteral router, no host name was
256 being passed; this would have caused $sender_host_name to have been used if
257 matching the list had actually called for a host name (not very likely,
258 since this list is usually IP addresses). A host name is now passed as
261 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
262 code that specifies a non-restarting handler (typically sigaction() in
263 modern systems) in an attempt to fix a rare and obscure crash bug.
265 25. Narrowed the window for a race in the daemon that could cause it to ignore
266 SIGCHLD signals. This is not a major problem, because they are used only to
267 wake it up if nothing else does.
269 26. A malformed maildirsize file could cause Exim to calculate negative values
270 for the mailbox size or file count. Odd effects could occur as a result.
271 The maildirsize information is now recalculated if the size or filecount
274 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
275 support for a long time. Removed HAVE_SYS_VFS_H.
277 28. Installed the latest version of exipick from John Jetmore.
279 29. In an address list, if the pattern was not a regular expression, an empty
280 subject address (from a bounce message) matched only if the pattern was an
281 empty string. Non-empty patterns were not even tested. This was the wrong
282 because it is perfectly reasonable to use an empty address as part of a
283 database query. An empty address is now tested by patterns that are
284 lookups. However, all the other forms of pattern expect the subject to
285 contain a local part and a domain, and therefore, for them, an empty
286 address still always fails if the pattern is not itself empty.
288 30. Exim went into a mad DNS loop when attempting to do a callout where the
289 host was specified on an smtp transport, and looking it up yielded more
292 31. Re-factored the code for checking spool and log partition space into a
293 function that finds that data and another that does the check. The former
294 is then used to implement four new variables: $spool_space, $log_space,
295 $spool_inodes, and $log_inodes.
297 32. The RFC2047 encoding function was originally intended for short strings
298 such as real names; it was not keeping to the 75-character limit for
299 encoded words that the RFC imposes. It now respects the limit, and
300 generates multiple encoded words if necessary. To be on the safe side, I
301 have increased the buffer size for the ${rfc2047: expansion operator from
304 33. It is now permitted to omit both strings after an "if" condition; if the
305 condition is true, the result is "true". As before, when the second string
306 is omitted, a false condition yields an empty string. This makes it less
307 cumbersome to write custom ACL and router conditions.
309 34. Failure to deliver a bounce message always caused it to be frozen, even if
310 there was an errors_to setting on the router. The errors_to setting is now
313 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
314 canonical form (fully expanded) before being placed in
315 $sender_host_address.
317 36. The table in the code that translates DNS record types into text (T_A to
318 "A" for instance) was missing entries for NS and CNAME. It is just possible
319 that this could have caused confusion if both these types were looked up
320 for the same domain, because the text type is used as part of Exim's
321 per-process caching. But the chance of anyone hitting this buglet seems
324 37. The dnsdb lookup has been extended in a number of ways.
326 (1) There is a new type, "zns", which walks up the domain tree until it
327 finds some nameserver records. It should be used with care.
329 (2) There is a new type, "mxh", which is like "mx" except that it returns
330 just the host names, not the priorities.
332 (3) It is now possible to give a list of domains (or IP addresses) to be
333 looked up. The behaviour when one of the lookups defers can be
334 controlled by a keyword.
336 (4) It is now possible to specify the separator character for use when
337 multiple records are returned.
339 38. The dnslists ACL condition has been extended: it is now possible to supply
340 a list of IP addresses and/or domains to be looked up in a particular DNS
343 39. Added log_selector=+queue_time_overall.
345 40. When running the queue in the test harness, wait just a tad after forking a
346 delivery process, to get repeatability of debugging output.
348 41. Include certificate and key file names in error message when GnuTLS fails
349 to set them up, because the GnuTLS error message doesn't include the name
350 of the failing file when there is a problem reading it.
352 42. Allow both -bf and -bF in the same test run.
354 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
356 44. The "Exiscan patch" is now merged into the mainline Exim source.
358 45. Sometimes the final signoff response after QUIT could fail to get
359 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
360 < 0 before doing a fflush(). This bug looks as though it goes back to the
361 introduction of TLS in release 3.20, but "sometimes" must have been rare
362 because the tests only now provoked it.
364 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
365 (this can affect the format of dates).
367 47. exim_tidydb, when checking for the continued existence of a message for
368 which it has found a message-specific retry record, was not finding
369 messages that were in split spool directories. Consequently, it was
370 deleting retry records that should have stayed in existence.
372 48. Steve fixed some bugs in eximstats.
374 49. The SPA authentication driver was not abandoning authentication and moving
375 on to the next authenticator when an expansion was forced to fail,
376 contradicting the general specification for all authenticators. Instead it
377 was generating a temporary error. It now behaves as specified.
379 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
380 (the order specifies the preference for clients). The order is now AES256,
381 AES128, 3DES, ARCFOUR128.
383 51. Small patch to Sieve code - explicitly set From: when generating an
386 52. Exim crashed if a remote delivery caused a very long error message to be
387 recorded - for instance if somebody sent an entire SpamAssassin report back
388 as a large number of 550 error lines. This bug was coincidentally fixed by
389 increasing the size of one of Exim's internal buffers (big_buffer) that
390 happened as part of the Exiscan merge. However, to be on the safe side, I
391 have made the code more robust (and fixed the comments that describe what
394 53. Now that there can be additional text after "Completed" in log lines (if
395 the queue_time_overall log selector is set), a one-byte patch to exigrep
396 was needed to allow it to recognize "Completed" as not the last thing in
399 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
400 patch that reportedly fixes this has been added. I am not expert enough to
401 create a test for it. This is what the patch creator wrote:
403 "I found a little strange behaviour of ldap code when working with
404 Windows 2003 AD Domain, where users was placed in more than one
405 Organization Units. When I tried to give exim partial DN, the exit code
406 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
407 But simultaneously result of request was absolutely normal ldap result,
408 so I produce this patch..."
410 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
411 REFERENCE, so I have modified the code to exclude the patch when that macro
414 55. Some experimental protocols are using DNS PTR records for new purposes. The
415 keys for these records are domain names, not reversed IP addresses. The
416 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
417 leaves it alone. Component reversal etc. now happens only for IP addresses.
420 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
422 57. Double the size of the debug message buffer (to 2048) so that more of very
423 long debug lines gets shown.
425 58. The exicyclog utility now does better if the number of log files to keep
426 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
428 59. Two changes related to the smtp_active_hostname option:
430 (1) $smtp_active_hostname is now available as a variable.
431 (2) The default for smtp_banner uses $smtp_active_hostname instead
432 of $primary_hostname.
434 60. The host_aton() function is supposed to be passed a string that is known
435 to be a valid IP address. However, in the case of IPv6 addresses, it was
436 not checking this. This is a hostage to fortune. Exim now panics and dies
437 if the condition is not met. A case was found where this could be provoked
438 from a dnsdb PTR lookup with an IPv6 address that had more than 8
439 components; fortuitously, this particular loophole had already been fixed
440 by change 4.50/55 above.
442 If there are any other similar loopholes, the new check in host_aton()
443 itself should stop them being exploited. The report I received stated that
444 data on the command line could provoke the exploit when Exim was running as
445 exim, but did not say which command line option was involved. All I could
446 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
450 61. There was a buffer overflow vulnerability in the SPA authentication code
451 (which came originally from the Samba project). I have added a test to the
452 spa_base64_to_bits() function which I hope fixes it.
455 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
456 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
458 63. The daemon start-up calls getloadavg() while still root for those OS that
459 need the first call to be done as root, but it missed one case: when
460 deliver_queue_load_max is set with deliver_drop_privilege. This is
461 necessary for the benefit of the queue runner, because there is no re-exec
462 when deliver_drop_privilege is set.
464 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
467 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
468 was in use, was not putting the data itself into the right store pool;
469 consequently, it could be overwritten for a subsequent message in the same
470 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
473 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
475 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
476 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
477 the function were treating the return as a boolean value, which happened to
478 work because 0=false and not-0=true, but is not correct code.
480 68. The host_aton() function was not handling scoped IPv6 addresses (those
481 with, for example, "%eth0" on the end) correctly.
483 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
484 negated items (that is, ~something) in unsigned ints. Some compilers
485 apparently mutter when there is no cast.
487 70. If an address verification called from an ACL failed, and did not produce a
488 user-specific message (i.e. there was only a "system" message), nothing was
489 put in $acl_verify_message. In this situation, it now puts the system
492 71. Change 4.23/11 added synchronization checking at the start of an SMTP
493 session; change 4.31/43 added the unwanted input to the log line - except
494 that it did not do this in the start of session case. It now does.
496 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
497 This is wrong and can cause the other end to generate a synchronization
498 error if it is another Exim or anything else that does the synchronization
499 check. A QUIT command is no longer sent after a timeout.
501 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
504 74. The defer_ok option of callout verification was not working if it was used
505 when verifying addresses in header lines, that is, for this case:
507 verify = header_sender/callout=defer_ok
509 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
510 those file descriptors could be used for SMTP connections. If anything
511 wrote to stderr (the example that came up was "warn" in embedded Perl), it
512 could be sent to the SMTP client, causing chaos. The daemon now opens
513 stdin, stdout, and stderr to /dev/null when it puts itself into the
516 76. Arrange for output from Perl's "warn" command to be written to Exim's main
517 log by default. The user can override this with suitable Perl magic.
519 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
520 the log message when discard triggers, was not working for the DATA ACL or
521 for the non-SMTP ACL.
523 78. Error message wording change in sieve.c.
525 79. If smtp_accept_max_per_host was set, the number of connections could be
526 restricted to fewer than expected, because the daemon was trying to set up
527 a new connection before checking whether the processes handling previous
528 connections had finished. The check for completed processes is now done
529 earlier. On busy systems, this bug wouldn't be noticed because something
530 else would have woken the daemon, and it would have reaped the completed
533 80. If a message was submitted locally by a user whose login name contained one
534 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
535 It caused a spool format error. I have fixed the spool reading code. A
536 related problem was that the "from" clause in the Received: line became
537 illegal because of the space(s). It is now covered by ${quote_local_part.
539 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
542 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
544 83. Patch from Sieve maintainer to fix unterminated string problem in
547 84. Some minor changes to the Linux configuration files to help with other
548 OS variants using glibc.
550 85. One more patch for Sieve to update vacation handling to latest spec.
553 ----------------------------------------------------
554 See the note above about the 4.44 and 4.50 releases.
555 ----------------------------------------------------
561 1. Change 4.43/35 introduced a bug that caused file counts to be
562 incorrectly computed when quota_filecount was set in an appendfile
565 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
566 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
569 3. Give more explanation in the error message when the command for a transport
570 filter fails to execute.
572 4. There are several places where Exim runs a non-Exim command in a
573 subprocess. The SIGUSR1 signal should be disabled for these processes. This
574 was being done only for the command run by the queryprogram router. It is
575 now done for all such subprocesses. The other cases are: ${run, transport
576 filters, and the commands run by the lmtp and pipe transports.
578 5. Some older OS have a limit of 256 on the maximum number of file
579 descriptors. Exim was using setrlimit() to set 1000 as a large value
580 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
581 systems. I've change it so that if it can't get 1000, it tries for 256.
583 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
584 was an oversight, and furthermore, ever since the addition of extra
585 controls (e.g. 4.43/32), the checks on when to allow different forms of
586 "control" were broken. There should now be diagnostics for all cases when a
587 control that does not make sense is encountered.
589 7. $recipients is now available in the predata ACL (oversight).
591 8. Tidy the search cache before the fork to do a delivery from a message
592 received from the command line. Otherwise the child will trigger a lookup
593 failure and thereby defer the delivery if it tries to use (for example) a
594 cached ldap connection that the parent has called unbind on.
596 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
597 of $address_data from the recipient verification was clobbered by the
600 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
601 was its contents. (It was OK if the option was not defined at all.)
603 11. A "Completed" log line is now written for messages that are removed from
604 the spool by the -Mrm option.
606 12. $host_address is now set to the target address during the checking of
609 13. When checking ignore_target_hosts for an ipliteral router, no host name was
610 being passed; this would have caused $sender_host_name to have been used if
611 matching the list had actually called for a host name (not very likely,
612 since this list is usually IP addresses). A host name is now passed as
615 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
616 code that specifies a non-restarting handler (typically sigaction() in
617 modern systems) in an attempt to fix a rare and obscure crash bug.
619 15. Narrowed the window for a race in the daemon that could cause it to ignore
620 SIGCHLD signals. This is not a major problem, because they are used only to
621 wake it up if nothing else does.
623 16. A malformed maildirsize file could cause Exim to calculate negative values
624 for the mailbox size or file count. Odd effects could occur as a result.
625 The maildirsize information is now recalculated if the size or filecount
628 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
629 support for a long time. Removed HAVE_SYS_VFS_H.
631 18. Updated exipick to current release from John Jetmore.
633 19. Allow an empty sender to be matched against a lookup in an address list.
634 Previously the only cases considered were a regular expression, or an
637 20. Exim went into a mad DNS lookup loop when doing a callout where the
638 host was specified on the transport, if the DNS lookup yielded more than
641 21. The RFC2047 encoding function was originally intended for short strings
642 such as real names; it was not keeping to the 75-character limit for
643 encoded words that the RFC imposes. It now respects the limit, and
644 generates multiple encoded words if necessary. To be on the safe side, I
645 have increased the buffer size for the ${rfc2047: expansion operator from
648 22. Failure to deliver a bounce message always caused it to be frozen, even if
649 there was an errors_to setting on the router. The errors_to setting is now
652 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
653 canonical form (fully expanded) before being placed in
654 $sender_host_address.
656 24. Updated eximstats to version 1.33
658 25. Include certificate and key file names in error message when GnuTLS fails
659 to set them up, because the GnuTLS error message doesn't include the name
660 of the failing file when there is a problem reading it.
662 26. Expand error message when OpenSSL has problems setting up cert/key files.
665 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
666 (this can affect the format of dates).
668 28. exim_tidydb, when checking for the continued existence of a message for
669 which it has found a message-specific retry record, was not finding
670 messages that were in split spool directories. Consequently, it was
671 deleting retry records that should have stayed in existence.
673 29. eximstats updated to version 1.35
674 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
675 1.35 - bugfix such that pie charts by volume are generated correctly
677 30. The SPA authentication driver was not abandoning authentication and moving
678 on to the next authenticator when an expansion was forced to fail,
679 contradicting the general specification for all authenticators. Instead it
680 was generating a temporary error. It now behaves as specified.
682 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
683 (the order specifies the preference for clients). The order is now AES256,
684 AES128, 3DES, ARCFOUR128.
686 31. Small patch to Sieve code - explicitly set From: when generating an
689 32. Exim crashed if a remote delivery caused a very long error message to be
690 recorded - for instance if somebody sent an entire SpamAssassin report back
691 as a large number of 550 error lines. This bug was coincidentally fixed by
692 increasing the size of one of Exim's internal buffers (big_buffer) that
693 happened as part of the Exiscan merge. However, to be on the safe side, I
694 have made the code more robust (and fixed the comments that describe what
697 33. Some experimental protocols are using DNS PTR records for new purposes. The
698 keys for these records are domain names, not reversed IP addresses. The
699 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
700 leaves it alone. Component reversal etc. now happens only for IP addresses.
703 34. The host_aton() function is supposed to be passed a string that is known
704 to be a valid IP address. However, in the case of IPv6 addresses, it was
705 not checking this. This is a hostage to fortune. Exim now panics and dies
706 if the condition is not met. A case was found where this could be provoked
707 from a dnsdb PTR lookup with an IPv6 address that had more than 8
708 components; fortuitously, this particular loophole had already been fixed
709 by change 4.50/55 or 4.44/33 above.
711 If there are any other similar loopholes, the new check in host_aton()
712 itself should stop them being exploited. The report I received stated that
713 data on the command line could provoke the exploit when Exim was running as
714 exim, but did not say which command line option was involved. All I could
715 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
719 35. There was a buffer overflow vulnerability in the SPA authentication code
720 (which came originally from the Samba project). I have added a test to the
721 spa_base64_to_bits() function which I hope fixes it.
724 36. The daemon start-up calls getloadavg() while still root for those OS that
725 need the first call to be done as root, but it missed one case: when
726 deliver_queue_load_max is set with deliver_drop_privilege. This is
727 necessary for the benefit of the queue runner, because there is no re-exec
728 when deliver_drop_privilege is set.
730 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
731 was in use, was not putting the data itself into the right store pool;
732 consequently, it could be overwritten for a subsequent message in the same
733 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
736 38. Sometimes the final signoff response after QUIT could fail to get
737 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
738 < 0 before doing a fflush(). This bug looks as though it goes back to the
739 introduction of TLS in release 3.20, but "sometimes" must have been rare
740 because the tests only now provoked it.
746 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
747 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
748 Now it returns an integer. A number of calls were still expecting a T/F
749 return. Fortuitously, in all cases, the tests worked in OK situations,
750 which is the norm. However, things would have gone wrong on any write
751 failures on the smtp file descriptor. This function is used when sending
752 messages over SMTP and also when doing verify callouts.
754 2. When Exim is called to do synchronous delivery of a locally submitted
755 message (the -odf or -odi options), it no longer closes stderr before doing
758 3. Implemented the mua_wrapper option.
760 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
762 5. Implemented the functions header_remove(), header_testname(),
763 header_add_at_position(), and receive_remove_recipient(), and exported them
766 6. If an ACL "warn" statement specified the addition of headers, Exim already
767 inserted X-ACL-Warn: at the start if there was no header name. However, it
768 was not making this test for the second and subsequent header lines if
769 there were newlines in the string. This meant that an invalid header could
770 be inserted if Exim was badly configured.
772 7. Allow an ACL "warn" statement to add header lines at the start or after all
773 the Received: headers, as well as at the end.
775 8. Added the rcpt_4xx retry error code.
777 9. Added postmaster_mailfrom=xxx to callout verification option.
779 10. Added mailfrom=xxxx to the callout verification option, for verify=
782 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
783 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
785 12. Inserted some casts to stop certain compilers warning when using pointer
786 differences as field lengths or precisions in printf-type calls (mostly
787 affecting debugging statements).
789 13. Added optional readline() support for -be (dynamically loaded).
791 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
792 same clock tick as a message's arrival, so that its received time was the
793 same as the "first fail" time on the retry record, and that message
794 remained on the queue past the ultimate address timeout, every queue runner
795 would try a delivery (because it was past the ultimate address timeout) but
796 after another failure, the ultimate address timeout, which should have then
797 bounced the address, did not kick in. This was a "< instead of <=" error;
798 in most cases the first failure would have been in the next clock tick
799 after the received time, and all would be well.
801 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
802 being recognized when the domain list was tested by the match_domain
803 condition in an expansion string.
805 16. Added the ${str2b64: operator.
807 17. Exim was always calling setrlimit() to set a large limit for the number of
808 processes, without checking whether the existing limit was already
809 adequate. (It did check for the limit on file descriptors.) Furthermore,
810 errors from getrlimit() and setrlimit() were being ignored. Now they are
811 logged to the main and panic logs, but Exim does carry on, to try to do its
812 job under whatever limits there are.
814 18. Imported PCRE 5.0.
816 19. Trivial typo in log message " temporarily refused connection" (the leading
819 20. If the log selector return_path_on_delivery was set and an address was
820 redirected to /dev/null, the delivery process crashed because it assumed
821 that a return path would always be set for a "successful" delivery. In this
822 case, the whole delivery is bypassed as an optimization, and therefore no
825 21. Internal re-arrangement: the function for sending a challenge and reading
826 a response while authentication was assuming a zero-terminated challenge
827 string. It's now changed to take a pointer and a length, to allow for
828 binary data in such strings.
830 22. Added the cyrus_sasl authenticator (code supplied by MBM).
832 23. Exim was not respecting finduser_retries when seeking the login of the
833 uid under which it was called; it was always trying 10 times. (The default
834 setting of finduser_retries is zero.) Also, it was sleeping after the final
835 failure, which is pointless.
837 24. Implemented tls_on_connect_ports.
839 25. Implemented acl_smtp_predata.
841 26. If the domain in control=submission is set empty, Exim assumes that the
842 authenticated id is a complete email address when it generates From: or
843 Sender: header lines.
845 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
846 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
847 chown and chgrp in /bin and hostname in /usr/bin.
849 28. Exim was keeping the "process log" file open after each use, just as it
850 does for the main log. This opens the possibility of it remaining open for
851 long periods when the USR1 signal hits a daemon. Occasional processlog
852 errors were reported, that could have been caused by this. Anyway, it seems
853 much more sensible not to leave this file open at all, so that is what now
856 29. The long-running daemon process does not normally write to the log once it
857 has entered its main loop, and it closes the log before doing so. This is
858 so that log files can straightforwardly be renamed and moved. However,
859 there are a couple of unusual error situations where the daemon does write
860 log entries, and I had neglected to close the log afterwards.
862 30. The text of an SMTP error response that was received during a remote
863 delivery was being truncated at 512 bytes. This is too short for some of
864 the long messages that one sometimes sees. I've increased the limit to
867 31. It is now possible to make retry rules that apply only when a message has a
868 specific sender, in particular, an empty sender.
870 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
871 it possible to be selective about when SMTP synchronization is enforced.
873 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
875 32. Implemented hosts_connection_nolog.
877 33. Added an ACL for QUIT.
879 34. Setting "delay_warning=" to disable warnings was not working; it gave a
882 35. Added mailbox_size and mailbox_filecount to appendfile.
884 36. Added control = no_multiline_responses to ACLs.
886 37. There was a bug in the logic of the code that waits for the clock to tick
887 in the case where the clock went backwards by a substantial amount such
888 that the microsecond fraction of "now" was more than the microsecond
889 fraction of "then" (but the whole seconds number was less).
891 38. Added support for the libradius Radius client library this is found on
892 FreeBSD (previously only the radiusclient library was supported).
898 1. When certain lookups returned multiple values in the form name=value, the
899 quoting of the values was not always being done properly. Specifically:
900 (a) If the value started with a double quote, but contained no whitespace,
902 (b) If the value contained whitespace other than a space character (i.e.
903 tabs or newlines or carriage returns) it was not quoted.
904 This fix has been applied to the mysql and pgsql lookups by writing a
905 separate quoting function and calling it from the lookup code. The fix
906 should probably also be applied to nisplus, ibase and oracle lookups, but
907 since I cannot test any of those, I have not disturbed their existing code.
909 2. A hit in the callout cache for a specific address caused a log line with no
910 reason for rejecting RCPT. Now it says "Previous (cached) callout
911 verification failure".
913 3. There was an off-by-one bug in the queryprogram router. An over-long
914 return line was truncated at 256 instead of 255 characters, thereby
915 overflowing its buffer with the terminating zero. As well as fixing this, I
916 have increased the buffer size to 1024 (and made a note to document this).
918 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
919 when Exim is waiting for an SMTP response from a remote server, Exim
920 restarts its select() call on the socket, thereby resetting its timeout.
921 This is not a problem when such interrupts are rare. Somebody set up a cron
922 job to run exiwhat every 2 minutes, which is less than the normal select()
923 timeout (5 or 10 minutes). This meant that the select() timeout never
924 kicked in because it was always reset. I have fixed this by comparing the
925 time when an interrupt arrives with the time at the start of the first call
926 to select(). If more time than the timeout has elapsed, the interrupt is
927 treated as a timeout.
929 5. Some internal re-factoring in preparation for the addition of Sieve
930 extensions (by MH). In particular, the "personal" test is moved to a
931 separate function, and given an option for scanning Cc: and Bcc: (which is
932 not set for Exim filters).
934 6. When Exim created an email address using the login of the caller as the
935 local part (e.g. when creating a From: or Sender: header line), it was not
936 quoting the local part when it contained special characters such as @.
938 7. Installed new OpenBSD configuration files.
940 8. Reworded some messages for syntax errors in "and" and "or" conditions to
941 try to make them clearer.
943 9. Callout options, other than the timeout value, were being ignored when
944 verifying sender addresses in header lines. For example, when using
946 verify = header_sender/callout=no_cache
948 the cache was (incorrectly) being used.
950 10. Added a missing instance of ${EXE} to the exim_install script; this affects
951 only the Cygwin environment.
953 11. When return_path_on_delivery was set as a log selector, if different remote
954 addresses in the same message used different return paths and parallel
955 remote delivery occurred, the wrong values would sometimes be logged.
956 (Whenever a remote delivery process finished, the return path value from
957 the most recently started remote delivery process was logged.)
959 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
960 lines when AUTH and/or TLS are in use. This is the "received protocol"
961 field. Exim used to use "asmtp" for authenticated SMTP, without any
962 indication (in the protocol name) for TLS use. Now it follows the RFC and
963 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
964 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
965 names appear in log lines as well as in Received: header lines.
967 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
968 extensions, and comparison tests, and to fix some bugs.
970 14. Changes to the "personal" filter test:
972 (1) The test was buggy in that it was just doing the equivalent of
973 "contains" tests on header lines. For example, if a user's address was
974 anne@some.where, the "personal" test would incorrectly be true for
976 To: susanne@some.where
978 This test is now done by extracting each address from the header in turn,
979 and checking the entire address. Other tests that are part of "personal"
980 are now done using regular expressions (for example, to check local parts
981 of addresses in From: header lines).
983 (2) The list of non-personal local parts in From: addresses has been
984 extended to include "listserv", "majordomo", "*-request", and "owner-*",
985 taken from the Sieve specification recommendations.
987 (3) If the message contains any header line starting with "List-" it is
988 treated as non-personal.
990 (4) The test for "circular" in the Subject: header line has been removed
991 because it now seems ill-conceived.
993 15. Minor typos in src/EDITME comments corrected.
995 16. Installed latest exipick from John Jetmore.
997 17. If headers_add on a router specified a text string that was too long for
998 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
999 of string_sprintf() is now avoided.
1001 18. $message_body_size was not set (it was always zero) when running the DATA
1002 ACL and the local_scan() function.
1004 19. For the "mail" command in an Exim filter, no default was being set for
1005 the once_repeat time, causing a random time value to be used if "once" was
1006 specified. (If the value happened to be <= 0, no repeat happened.) The
1007 default is now 0s, meaning "never repeat". The "vacation" command was OK
1008 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1009 (I found it when inspecting the code).
1011 20. There is now an overall timeout for performing a callout verification. It
1012 defaults to 4 times the callout timeout, which applies to individual SMTP
1013 commands during the callout. The overall timeout applies when there is more
1014 than one host that can be tried. The timeout is checked before trying the
1015 next host. This prevents very long delays if there are a large number of
1016 hosts and all are timing out (e.g. when the network connections are timing
1017 out). The value of the overall timeout can be changed by specifying an
1018 additional sub-option for "callout", called "maxwait". For example:
1020 verify = sender/callout=5s,maxwait=20s
1022 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1023 to the end before writing, but this should make it even safer).
1025 22. Exim was forgetting that it had advertised PIPELINING for the second and
1026 subsequent messages on an SMTP connection. It was also not resetting its
1027 memory on STARTTLS and an internal HELO.
1029 23. When Exim logs an SMTP synchronization error within a session, it now
1030 records whether PIPELINING has been advertised or not.
1032 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1033 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1034 rather than long int.
1036 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1038 26. Added the never_mail option to autoreply.
1044 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1045 crash if the getsockname() call failed; this can happen if a connection is
1046 closed very soon after it is established. The problem was simply in the
1047 order in which certain operations were done, causing Exim to try to write
1048 to the SMTP stream before it had set up the file descriptor. The bug has
1049 been fixed by making things happen in the correct order.
1055 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1056 before the connection was closed, thus losing the rejection response.
1058 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1059 some early Solaris releases, but causes trouble in current releases where
1060 socklen_t is defined.
1062 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1065 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1068 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1069 that releasing the top of it at the end releases what was used for sub-
1070 expansions (unless the block got too big). However, discard this block if
1071 the first thing is a variable or header, so that we can use its block when
1072 it is dynamic (useful for very large $message_headers, for example).
1074 6. Lookups now cache *every* query, not just the most recent. A new, separate
1075 store pool is used for this. It can be recovered when all lookup caches are
1076 flushed. Lookups now release memory at the end of their result strings.
1077 This has involved some general refactoring of the lookup sources.
1079 7. Some code has been added to the store_xxx() functions to reduce the amount
1080 of flapping under certain conditions.
1082 8. log_incoming_interface used to affect only the <= reception log lines. Now
1083 it causes the local interface and port to be added to several more SMTP log
1084 lines, for example "SMTP connection from", and rejection lines.
1086 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1088 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1090 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1091 could be overwritten at the end of the current message (or the start of a
1092 new message if it was set in a HELO ACL). The value is now preserved for
1093 the duration of the SMTP connection.
1095 12. If a transport had a headers_rewrite setting, and a matching header line
1096 contained an unqualified address, that address was qualified, even if it
1097 did not match any rewriting rules. The underlying bug was that the values
1098 of the flags that permit the existence of unqualified sender and recipient
1099 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1100 non-local messages, and by -bnq for local messages) were not being
1101 preserved with the message after it was received.
1103 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1104 "next input=" as part of the text comprising the host identity instead of
1105 the correct text. The code was using the same buffer for two different
1106 strings. However, depending on which order the printing function evaluated
1107 its arguments, the bug did not always show up. Under Linux, for example, my
1108 test suite worked just fine.
1110 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1111 doesn't work with some older versions of Perl. It has been changed to "my",
1112 which in any case is probably the better facility to use.
1114 15. A really picky compiler found some instances of statements for creating
1115 error messages that either had too many or two few arguments for the format
1118 16. The size of the buffer for calls to the DNS resolver has been increased
1119 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1120 for addresses that have a lot of PTR records. This alleviates a problem; it
1121 does not fully solve it.
1123 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1124 buffer now truncates the list and logs the incident, which is the same
1125 action as happens when Exim is looking up a host name and its aliases.
1126 Previously in this situation something unpredictable would happen;
1127 sometimes it was "internal error: store_reset failed".
1129 18. If a server dropped the connection unexpectedly when an Exim client was
1130 using GnuTLS and trying to read a response, the client delivery process
1131 crashed while trying to generate an error log message.
1133 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1134 string, for example:
1136 warn message = H1: something\nH2: something
1138 the text was added as a single header line from Exim's point of view
1139 though it ended up OK in the delivered message. However, searching for the
1140 second and subsequent header lines using $h_h2: did not work. This has been
1141 fixed. Similarly, if a system filter added multiple headers in this way,
1142 the routers could not see them.
1144 20. Expanded the error message when iplsearch is called with an invalid key to
1145 suggest using net-iplsearch in a host list.
1147 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1148 ACLs are no longer actually imposed (and a message to that effect is
1151 22. If a "gecos" field in a passwd entry contained escaped characters, in
1152 particular, if it contained a \" sequence, Exim got it wrong when building
1153 a From: or a Sender: header from that name. A second bug also caused
1154 incorrect handling when an unquoted " was present following a character
1155 that needed quoting.
1157 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1158 was not being matched caselessly.
1160 24. Arranged for all hyphens in the exim.8 source to be escaped with
1163 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1164 properly. Recipient callout cache records were still being keyed to include
1165 the sender, even when use_sender was set false. This led to far more
1166 callouts that were necessary. The sender is no longer included in the key
1167 when use_sender is false.
1169 26. Added "control = submission" modifier to ACLs.
1171 27. Added the ${base62d: operator to decode base 62 numbers.
1173 28. dnsdb lookups can now access SRV records.
1175 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1176 the configuration file.
1178 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1179 (-v) mode. This makes the output for a verbose queue run more intelligible.
1181 31. Added a use_postmaster feature to recipient callouts.
1183 32. Added the $body_zerocount variable, containing the number of binary zero
1184 bytes in the message body.
1186 33. The time of last modification of the "new" subdirectory is now used as the
1187 "mailbox time last read" when there is a quota error for a maildir
1190 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1192 35. Added +ignore_unknown as a special item in host lists.
1194 36. Code for decoding IPv6 addresses in host lists is now included, even if
1195 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1196 address was recognized as an IP address, but was then not correctly decoded
1197 into binary, causing unexpected and incorrect effects when compared with
1204 1. Very minor rewording of debugging text in manualroute to say "list of
1205 hosts" instead of "hostlist".
1207 2. If verify=header_syntax was set, and a header line with an unqualified
1208 address (no domain) and a large number of spaces between the end of the
1209 name and the colon was received, the reception process suffered a buffer
1210 overflow, and (when I tested it) crashed. This was caused by some obsolete
1211 code that should have been removed. The fix is to remove it!
1213 3. When running in the test harness, delay a bit after writing a bounce
1214 message to get a bit more predictability in the log output.
1216 4. Added a call to search_tidyup() just before forking a reception process. In
1217 theory, someone could use a lookup in the expansion of smtp_accept_max_
1218 per_host which, without the tidyup, could leave open a database connection.
1220 5. Added the variables $recipient_data and $sender_data which get set from a
1221 lookup success in an ACL "recipients" or "senders" condition, or a router
1222 "senders" option, similar to $domain_data and $local_part_data.
1224 6. Moved the writing of debug_print from before to after the "senders" test
1227 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1228 problems for message scanning, either using a data ACL, or using
1229 local_scan() because the Received: header was not generated till after they
1230 were called (in order to set the time as the time of reception completion).
1231 I have revised the way this works. The header is now generated after the
1232 body is received, but before the ACL or local_scan() are called. After they
1233 are run, the timestamp in the header is updated.
1239 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1240 before starting a queue runner without re-exec. This happened only when
1241 deliver_drop_privilege was set or when the Exim user was set to root. The
1242 effect of the bug was that timeouts during subsequent deliveries caused
1243 crashes instead of being properly handled. The handler is now left at its
1244 default (and expected) setting.
1246 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1247 message, again when deliver_drop_privilege is set or Exim is run as root.
1248 The bug described in (1) was not present in this case, but the tidying up
1249 of the other signals was missing. I have made the two cases consistent.
1251 3. The ignore_target_hosts setting on a manualroute router was being ignored
1252 for hosts that were looked up using the /MX notation.
1254 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1257 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1258 operated on the sender address. After changing the $sender_address to <>
1259 for the sender address verify, Exim was re-instated it as the original
1260 (before rewriting) address, but remembering that it had rewritten it, so it
1261 wasn't rewriting it again. This bug also had the effect of breaking the
1262 sender address verification caching when the sender address was rewritten.
1264 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1265 This has been changed so that if the ip literal address matches
1266 ignore_target_hosts, the router declines.
1268 7. Added expansion conditions match_domain, match_address, and match_local_
1269 part (NOT match_host).
1271 8. The placeholder for the Received: header didn't have a length field set.
1273 9. Added code to Exim itself and to exim_lock to test for a specific race
1274 condition that could lead to file corruption when using MBX delivery. The
1275 issue is with the lockfile that is created in /tmp. If this file is removed
1276 after a process has opened it but before that process has acquired a lock,
1277 there is the potential for a second process to recreate the file and also
1278 acquire a lock. This could lead to two Exim processes writing to the file
1279 at the same time. The added code performs the same test as UW imapd; it
1280 checks after acquiring the lock that its file descriptor still refers to
1281 the same named file.
1283 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1284 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1287 11. Added the smtp_active_hostname option. If used, this will typically be made
1288 to depend on the incoming interface address. Because $interface_address is
1289 not set up until the daemon has forked a reception process, error responses
1290 that can happen earlier (such as "too many connections") no longer contain
1293 12. If an expansion in a condition on a "warn" statement fails because a lookup
1294 defers, the "warn" statement is abandoned, and the next ACL statement is
1295 processed. Previously this caused the whole ACL to be aborted.
1297 13. Added the iplsearch lookup type.
1299 14. Added ident_timeout as a log selector.
1301 15. Added tls_certificate_verified as a log selector.
1303 16. Added a global option tls_require_ciphers (compare the smtp transport
1304 option of the same name). This controls incoming TLS connections.
1306 17. I finally figured out how to make tls_require_ciphers do a similar thing
1307 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1308 before starting the TLS session.
1310 18. Tabs are now shown as \t in -bP output.
1312 19. If the log selector return_path_on_delivery was set, Exim crashed when
1313 bouncing a message because it had too many Received: header lines.
1315 20. If two routers both had headers_remove settings, and the first one included
1316 a superfluous trailing colon, the final name in the first list and the
1317 first name in the second list were incorrectly joined into one item (with a
1318 colon in the middle).
1324 1. Added -C and -D options to the exinext utility, mainly to make it easier
1325 to include in the automated testing, but these could be helpful when
1326 multiple configurations are in use.
1328 2. The exinext utility was not formatting the output nicely when there was
1329 an alternate port involved in the retry record key, nor when there was a
1330 message id as well (for retries that were specific to a specific message
1331 and a specific host). It was also confused by IPv6 addresses, because of
1332 the additional colons they contain. I have fixed the IPv4 problem, and
1333 patched it up to do a reasonable job for IPv6.
1335 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1336 delivery, the log line now contains "pipelined" if PIPELINING was used.
1338 4. An SMTP transport process used to panic and die if the bind() call to set
1339 an explicit outgoing interface failed. This has been changed; it is now
1340 treated in the same way as a connect() failure.
1342 5. A reference to $sender_host_name in the part of a conditional expansion
1343 that was being skipped was still causing a DNS lookup. This no longer
1346 6. The def: expansion condition was not recognizing references to header lines
1347 that used bh_ and bheader_.
1349 7. Added the _cache feature to named lists.
1351 8. The code for checking quota_filecount in the appendfile transport was
1352 allowing one more file than it should have been.
1354 9. For compatibility with Sendmail, the command line option
1362 and sets the incoming protocol and host name (for trusted callers). The
1363 host name and its colon can be omitted when only the protocol is to be set.
1364 Note the Exim already has two private options, -pd and -ps, that refer to
1365 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1366 "s", but I don't think that's a major issue.
1368 10. A number of refactoring changes to the code, none of which should affect
1371 (a) The number of logging options was getting close to filling up the
1372 32-bit word that was used as a bit map. I have split them into two classes:
1373 those that are passed in the argument to log_write(), and those that are
1374 only ever tested independently outside of that function. These are now in
1375 separate 32-bit words, so there is plenty of room for expansion again.
1376 There is no change in the user interface or the logging behaviour.
1378 (b) When building, for example, log lines, the code previously used a
1379 macro that called string_cat() twice, in order to add two strings. This is
1380 not really sufficiently general. Furthermore, there was one instance where
1381 it was actually wrong because one of the argument was used twice, and in
1382 one call a function was used. (As it happened, calling the function twice
1383 did not affect the overall behaviour.) The macro has been replaced by a
1384 function that can join an arbitrary number of extra strings onto a growing
1387 (c) The code for expansion conditions now uses a table and a binary chop
1388 instead of a serial search (which was left over from when there were very
1389 few conditions). Also, it now recognizes conditions like "pam" even when
1390 the relevant support is not compiled in: a suitably worded error message is
1391 given if an attempt is made to use such a condition.
1393 11. Added ${time_interval:xxxxx}.
1395 12. A bug was causing one of the ddress fields not to be passed back correctly
1396 from remote delivery subprocesses. The field in question was not being
1397 subsequently used, so this caused to problems in practice.
1399 13. Added new log selectors queue_time and deliver_time.
1401 14. Might have fixed a bug in maildirsizefile handling that threw up
1402 "unexpected character" debug warnings, and recalculated the data
1403 unnecessarily. In any case, I expanded the warning message to give more
1406 15. Added the message "Restricted characters in address" to the statements in
1407 the default ACL that block characters like @ and % in local parts.
1409 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1410 Three changes have been made:
1412 (a) There was a serious bug; a negative response to MAIL caused the whole
1413 recipient domain to be cached as invalid, thereby blocking all messages
1414 to all local parts at the same domain, from all senders. This bug has
1415 been fixed. The domain is no longer cached after a negative response to
1416 MAIL if the sender used is not empty.
1418 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1421 (c) A new callout option, "use_sender" has been added for people who want
1422 the modified behaviour.
1428 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1431 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1432 indeed breaks things for older releases.
1434 3. Added additional logging to the case where there is a problem reading data
1435 from a filter that is running in a subprocess using a pipe, in order to
1436 try to track down a specific problem.
1438 4. Testing facility fudge: when running in the test harness and attempting
1439 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1440 getting "No route to host". Convert this to a timeout.
1442 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1445 6. Some OS don't have socklen_t but use size_t instead. This affects the
1446 fifth argument of getsockopt() amongst other things. This is now
1447 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1448 can be set for individual OS. I have set it for SunOS5, OSF1, and
1449 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1450 some earlier ones do not.
1452 7. Change 4.30/15 was not doing the test caselessly.
1454 8. The standard form for an IPv6 address literal was being rejected by address
1455 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1456 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1457 this, as well as the form without the "IPv6" on the front (but only when
1458 address literals are enabled, of course).
1460 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1462 10. Exim crashed if a message with an empty sender address specified by -f
1463 encountered a router with an errors_to setting. This could be provoked only
1464 by a command such as
1468 where an empty string was supplied; "<>" did not hit this bug.
1470 11. Installed PCRE release 4.5.
1472 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1473 remained set. It is now erased.
1475 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1476 times from message ids (which are base 36 rather than the normal 62).
1478 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1479 were being counted as actual protocol errors, and logged if the log
1480 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1481 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1482 and DATA following a set of rejected RCPTs do not count as protocol errors.
1483 In other words, Exim assumes they were pipelined, though this may not
1484 actually be the case. Of course, in all cases the client gets an
1485 appropriate error code.
1487 15. If a lookup fails in an ACL condition, a message about the failure may
1488 be available; it is used if testing the ACL cannot continue, because most
1489 such messages specify what the cause of the deferral is. However, some
1490 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1491 that caused an old message to be retained and used if a later statement
1492 caused a defer, replacing the real cause of the deferral.
1494 16. If an IP address had so many PTR records that the DNS lookup buffer
1495 was not large enough to hold them, Exim could crash while trying to process
1496 the truncated data. It now detects and logs this case.
1498 17. Further to 4.21/58, another change has been made: if (and only if) the
1499 first line of a message (the first header line) ends with CRLF, a bare LF
1500 in a subsequent header line has a space inserted after it, so as not to
1501 terminate the header.
1503 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1504 unnecessarily, used atoi() instead of strtol(), and didn't check the
1505 termination when getting file sizes from file names by regex.
1507 19. Completely re-implemented the support for maildirsize files, in the light
1508 of a number of problems with the previous contributed implementation
1509 (4.30/29). In particular:
1511 . If the quota is zero, the maildirsize file is maintained, but no quota is
1514 . If the maildir directory does not exist, it is created before any attempt
1515 to write a maildirsize file.
1517 . The quota value in the file is just a cache; if the quota is changed in
1518 the transport, the new value overrides.
1520 . A regular expression is available for excluding directories from the
1523 20. The autoreply transport checks the characters in options that define the
1524 message's headers; it allows continued headers, but it was checking with
1525 isspace() after an embedded newline instead of explicitly looking for a
1528 21. If all the "regular" hosts to which an address was routed had passed their
1529 expiry times, and had not reached their retry times, the address was
1530 bounced, even if fallback hosts were defined. Now Exim should go on to try
1533 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1534 equivalent code in the SMTP transport. Some hosts send humungous responses
1535 to HELO/EHLO, more than 1024 it seems.
1537 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1538 gives compiler warnings in some environments. I've now done it "properly",
1541 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1542 (because of problems with the built-in one) was declared to return uschar *
1543 instead of char *, causing compiler failure.
1545 25. Fixed a file descriptor leak when processing alias/forward files.
1547 26. Fixed a minor format string issue in dbfn.c.
1549 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1551 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1552 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1555 29. When a sender address is verified, it is cached, to save repeating the test
1556 when there is more than one recipient in a message. However, when the
1557 verification involves a callout, it is possible for different callout
1558 options to be set for different recipients. It is too complicated to keep
1559 track of this in the cache, so now Exim always runs a verification when a
1560 callout is required, relying on the callout cache for the optimization.
1561 The overhead is duplication of the address routing, but this should not be
1564 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1565 to be verified with callout=postmaster, and the main callout worked but the
1566 postmaster check failed, the verification correctly failed. However, if a
1567 subsequent RCPT command asked for sender verification *without* the
1568 postmaster check, incorrect caching caused this verification also to fail,
1571 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1572 it was not caching the DNS options (qualify_single, search_parents) that
1573 were used when the lookup failed. A subsequent lookup with different
1574 options therefore always gave the same answer, though there were cases
1575 where it should not have. (Example: a "domains = !$mx_any" option on a
1576 dnslookup router: the "domains" option is always processed without any
1577 widening, but the router might have qualify_single set.) Now Exim uses the
1578 cached value only when the same options are set.
1580 32. Added John Jetmore's "exipick" utility to the distribution.
1582 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1583 than a timeout (e.g. a certificate is required, and is not provided), an
1584 Exim server now closes the connection immediately. Previously it waited for
1585 the client to close - but if the client is SSL, it seems that they each
1586 wait for each other, leading to a delay before one of them times out.
1588 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1589 maintained 0.8.x compatibility because I don't think many are using it, and
1590 it is clearly obsolete.
1592 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1595 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1596 client certificate was expired. A simple patch fixes this, though I don't
1597 understand the full logic of why the verify callback is called multiple
1600 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1601 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1602 which causes problems with some clients (such as the Certicom SSL Plus
1603 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1604 disables the coutermeasure allowing Eudora to connect."
1606 38. Exim was not checking that a write() to a log file succeeded. This could
1607 lead to Bad Things if a log got too big, in particular if it hit a file
1608 size limit. Exim now panics and dies if it cannot write to a log file, just
1609 as it does if it cannot open a log file.
1611 39. Modified OS/Makefile-Linux so that it now contains
1613 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1615 The two -D definitions ensure that Exim is compiled with large file
1616 support, which makes it possible to handle log files that are bigger than
1619 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1620 instance) a domain was checked against a named list that involved a lookup,
1621 causing $domain_data to be set, then another domain was checked against the
1622 same list, then the first domain was re-checked, the value of $domain_data
1623 after the final check could be wrong. In particular, if the second check
1624 failed, it could be set empty. This bug probably also applied to
1627 41. The strip_trailing_dot option was not being applied to the address given
1628 with the -f command-line option.
1630 42. The code for reading a message's header from the spool was incrementing
1631 $received_count, but never initializing it. This meant that the value was
1632 incorrect (doubled) while delivering a message in the same process in which
1633 it was received. In the most common configuration of Exim, this never
1634 happens - a fresh exec is done - but it can happen when
1635 deliver_drop_privilege is set.
1637 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1638 it now includes up to 150 characters of the unexpected data in the log
1641 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1642 and building data strings. The size of both of these buffers was 10 000
1643 bytes - far larger than anybody would *ever* want, thought I. Needless to
1644 say, somebody hit the limit. I have increased the maximum line length to
1645 20 000 and the maximum data length of concatenated lines to 100 000. I have
1646 also fixed two bugs, because there was no checking on these buffers. Tsk,
1647 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1648 buffer is too small.
1650 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1651 lsearch lookups. Now it does.
1653 46. When parsing a route_list item in a manualroute router, a fixed-length
1654 buffer was used for the list of hosts. I made this 1024 bytes long,
1655 thinking that nobody would ever have a list of hosts that long. Wrong.
1656 Somebody had a whole pile of complicated expansion conditions, and the
1657 string was silently truncated, leading to an expansion error. It turns out
1658 that it is easier to change to an unlimited length (owing to other changes
1659 that have happened since this code was originally written) than to build
1660 structure for giving a limitation error. The length of the item that
1661 expands into the list of hosts is now unlimited.
1663 47. The lsearch lookup could not handle data where the length of text line was
1664 more than 4095 characters. Such lines were truncated, leading to shortened
1665 data being returned. It should now handle lines of any length.
1667 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1668 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1671 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1672 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1673 no longer happens in comment lines. A list of replacements is now placed
1674 at the head of all of the source files, except those whose only change is
1675 to replace PERL_COMMAND in the very first #! line.
1677 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1678 messages on the queue, with a bottom-up merge sort, using code contributed
1679 by Michael Haardt. This should make operations like -bp somewhat faster on
1680 large queues. It won't affect queue runners, except when queue_run_in_order
1683 51. Installed eximstats 1.31 in the distribution.
1685 52. Added support for SRV lookups to the dnslookup router.
1687 53. If an ACL referred to $message_body or $message_body_end, the value was not
1688 reset for any messages that followed in the same SMTP session.
1690 54. The store-handling optimization for building very long strings was not
1691 differentiating between the different store pools. I don't think this
1692 actually made any difference in practice, but I've tidied it.
1694 55. While running the routers to verify a sender address, $sender_address
1695 was still set to the sender address. This is wrong, because when routing to
1696 send a bounce to the sender, it would be empty. Therefore, I have changed
1697 it so that, while verifying a sender address, $sender_address is set to <>.
1698 (There is no change to what happens when verifying a recipient address.)
1700 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1701 target A or AAAA records (if not already returned) without resetting the
1702 qualify_single or search_parents options of the DNS resolver. These are
1703 inappropriate in this case because the targets of MX and SRV records must
1704 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1705 target that, when qualified, matched something in the local domain. These
1706 two options are now turned off when doing these lookups.
1708 57. It seems that at least some releases of Reiserfs (which does not have the
1709 concept of a fixed number of inodes) returns zero and not -1 for the
1710 number of available inodes. This interacted badly with check_spool_inodes,
1711 which assumed that -1 was the "no such thing" setting. What I have done is
1712 to check that the total number of inodes is greater than zero before doing
1713 the test of how many are available.
1715 58. When a "warn" ACL statement has a log_message modifier, the message is
1716 remembered, and not repeated. This is to avoid a lot of repetition when a
1717 message has many recipients that cause the same warning to be written.
1718 Howewer, Exim was preserving the list of already written lines for an
1719 entire SMTP session, which doesn't seem right. The memory is now reset if a
1720 new message is started.
1722 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1723 debugging output unless log_rewrite was also set.
1725 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1726 of (int)(handle) when we know that handle contains (void *)(-1).
1728 61. The Exim daemon panic-logs an error return when it closes the incoming
1729 connection. However "connection reset by peer" seems to be common, and
1730 isn't really an error worthy of noting specially, so that particular error
1733 62. When Exim is trying to find all the local interfaces, it used to panic and
1734 die if the ioctl to get the interface flags failed. However, it seems that
1735 on at least one OS (Solaris 9) it is possible to have an interface that is
1736 included in the list of interfaces, but for which you get a failure error
1737 for this call. This happens when the interface is not "plumbed" into a
1738 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1739 failure of the "get flags" call assumes that the interface is down.
1741 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1742 makes life easier for people who are doing arithmetic on fields extracted
1743 from dates, where you often get leading zeros that should not be
1744 interpreted as octal.
1746 64. Added qualify_domain to the redirect router, to override the global
1749 65. If a pathologically long header line contained very many addresses (the
1750 report of this problem mentioned 10 000) and each of them was rewritten,
1751 Exim could use up a very large amount of memory. (It kept on making new
1752 copies of the header line as it rewrote, and never released the old ones.)
1753 At the expense of a bit more processing, the header rewriting function has
1754 been changed so that it no longer eats memory in this way.
1756 66. The generation of the Received: header has been moved from the time that a
1757 message starts to be received, to the time that it finishes. The timestamp
1758 in the Received: header should now be very close to that of the <= log
1759 line. There are two side-effects of this change:
1761 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1762 logged header lines no longer include the local Received: line, because
1763 it has not yet been created. The same applies to a copy of the message
1764 that is returned to a non-SMTP sender when a message is rejected.
1766 (b) When a filter file is tested using -bf, no additional Received: header
1767 is added to the test message. After some thought, I decided that this
1770 This change does not affect the value of $received_for. It is still set
1771 after address rewriting, but before local_scan() is called.
1773 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1775 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1776 gave an unhelpful panic error message, and a defer error. I have managed to
1777 change this behaviour so that it now rejects any supplied certificate,
1778 which seems right, as the list of acceptable certificates is empty.
1780 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1781 gave an unhelpful defer error. I have not managed to make this reject any
1782 supplied certificates, but the error message it gives is "no certificate
1783 supplied", which is not helpful.
1785 70. exigrep's output now also includes lines that are not associated with any
1786 message, but which match the given pattern. Implemented by a patch from
1787 Martin Sluka, which also tidied up the Perl a bit.
1789 71. Recipient callout verification, like sender verification, was using <> in
1790 the MAIL FROM command. This isn't really the right thing, since the actual
1791 sender may affect whether the remote host accepts the recipient or not. I
1792 have changed it to use the actual sender in the callout; this means that
1793 the cache record is now keyed on a recipient/sender pair, not just the
1794 recipient address. There doesn't seem to be a real danger of callout loops,
1795 since a callout by the remote host to check the sender would use <>.
1796 [SEE ABOVE: changed after hitting problems.]
1798 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1799 temporary errors. However, in the case of such a code being given after
1800 the end of a data transmission (i.e. after ".") Exim was failing to write
1801 a retry record for the message. (Yes, there was some broken host that was
1802 actually sending 8xx at this point.)
1804 73. An unknown lookup type in a host list could cause Exim to panic-die when
1805 the list was checked. (An example that provoked this was putting <; in the
1806 middle of a list instead of at the start.) If this happened during a DATA
1807 ACL check, a -D file could be left lying around. This kind of configuration
1808 error no longer causes Exim to die; instead it causes a defer errror. The
1809 incident is still logged to the main and panic logs.
1811 74. Buglet left over from Exim 3 conversion. The message "too many messages
1812 in one connection" was written to the rejectlog but not the mainlog, except
1813 when address rewriting (yes!) was being logged.
1815 75. Added write_rejectlog option.
1817 76. When a system filter was run not as root (that is, when system_filter_user
1818 was set), the values of the $n variables were not being returned to the
1819 main process; thus, they were not subsequently available in the $sn
1822 77. Added +return_path_on_delivery log selector.
1824 78. A connection timeout was being treated differently from recipients deferred
1825 when testing hosts_max_try with a message that was older than the host's
1826 retry timeout. (The host should not be counted, thus allowing all hosts to
1827 be tried at least once before bouncing.) This may have been the cause of an
1828 occasionally reported bug whereby a message would remain on the queue
1829 longer than the retry timeout, but would be bounced if a delivery was
1830 forced. I say "may" because I never totally pinned down the problem;
1831 setting up timeout/retry tests is difficult. See also the next item.
1833 79. The ultimate address timeout was not being applied to errors that involved
1834 a combination of host plus message (for example, a timeout on a MAIL
1835 command). When an address resolved to a number of possible hosts, and they
1836 were not all tried for each delivery (e.g. because of hosts_max_try), a
1837 message could remain on the queue longer than the retry timeout.
1839 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1842 81. Fixed an obscure SMTP outgoing bug which required at least the following
1843 conditions: (a) there was another message waiting for the same server;
1844 (b) the server returned 5xx to all RCPT commands in the first message so
1845 that the message was not completed; (c) the server dropped the connection
1846 or gave a negative response to the RSET that Exim sends to abort the
1847 transaction. The observed case was a dropped connection after DATA that had
1848 been sent in pipelining mode. That is, the server had advertised PIPELINING
1849 but was not implementing it correctly. The effect of the bug was incorrect
1850 behaviour, such as trying another host, and this could lead to a crash.
1856 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1857 and daemon.c were passed as pointers to ints; they should have been
1858 pointers to socklen_t variables (which are typically unsigned ints).
1860 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1863 3. Fixed a really odd bug that affected only the testing scheme; patching a
1864 certain fixed string in the binary changed the value of another string that
1865 happened to be identical to the end of the original first string.
1867 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1868 name", it returns that address as the IP address. On some operating
1869 systems (e.g. Solaris), it also passes back the IP address string as the
1870 "host name". However, on others (e.g. Linux), it passes back an empty
1871 string. Exim wasn't checking for this, and was changing the host name to an
1872 empty string, assuming it had been canonicized.
1874 5. Although rare, it is permitted to have more than one PTR record for a given
1875 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1876 all the names associated with an address, because they do in Solaris.
1877 However, it seems that they do not in Linux for data that comes from the
1878 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1879 I found this out when I moved to a new Linux workstation and tried to run
1880 the Exim test suite.
1882 To get round this problem I have changed the code so that it now does its
1883 own call to the DNS to look up PTR records when searching for a host name.
1884 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1885 addresses that are only in /etc/hosts are still found.
1887 This behaviour is, however, controlled by an option called host_lookup_
1888 order, which defaults to "bydns:byaddr". If people want to use the other
1889 order, or indeed, just use one or the other means of lookup, they can
1890 specify it in this variable.
1892 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1893 some operating systems, this comes back from gethostbyaddr() as an empty
1894 string, and this is what Exim used to test for. However, it seems that in
1895 other systems, "." is yielded. Exim now tests for this case too.
1897 7. The values of check_spool_space and check_log_space are now held internally
1898 as a number of kilobytes instead of an absolute number of bytes. If a
1899 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1900 kilobyte. This means that much larger values can be stored.
1902 8. Exim monitor: an attempt to get the action menu when not actually pointing
1903 at a message produces an empty menu entitled "No message selected". This
1904 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1905 no entries in it ("Shell widget menu has zero width and/or height"). So I
1906 have added a single, blank menu entry in this case.
1908 9. Added ${quote_local_part.
1910 10. MIME decoding is now applied to the contents of Subject: header lines when
1913 11. Now that a reference to $sender_host_address automatically causes a reverse
1914 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1915 host lookup before query-style lookups in lists that might use this
1916 variable. This has therefore been abolished, and the "net-" prefix is no
1917 longer necessary for query-style lookups.
1919 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1920 have been a typo for LFLAGS, so it has been changed.
1922 13. The install script calls Exim with "-C /dev/null" in order to find the
1923 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1924 to be output. Howeve, since Exim outputs its version number before the
1925 error, it didn't break the script. It just looked ugly. I fixed this by
1926 always allowing "-C /dev/null" if the caller is root.
1928 14. Ignore overlarge ACL variable number when reading spool file - insurance
1929 against a later release with more variables having written the file.
1931 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1932 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1933 this, as well as the form without the "IPv6" on the front.
1935 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1936 OS/Makefile-Darwin file.
1938 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1939 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1941 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1942 variable. However, some OS use ints for inodes. I've added cast to long int
1943 to get rid of the compiler warning.
1945 19. I had forgotten to lock out "/../" in configuration file names when
1946 ALT_CONFIG_PREFIX was set.
1948 20. Routers used for verification do not need to specify transports. However,
1949 if such a router generated a host list, and callout was configured, Exim
1950 crashed, because it could not find a port number from the (non-existent)
1951 transport. It now assumes port 25 in this circumstance.
1953 21. Added the -t option to exigrep.
1955 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1956 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1957 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1958 them is set, LOOKUP_LSEARCH is forced.
1960 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1962 24. Added sender and host information to the "rejected by local_scan()" log
1963 line; previously there was no indication of these.
1965 25. Added .include_if_exists.
1967 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1968 Linux. It turns out that not all file systems support this. Apparently some
1969 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1970 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1971 not supported on the file descriptor, is now ignored when Exim is trying to
1972 sync a directory. This applies only to Linux.
1974 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1976 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1977 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1980 29. Added support for maildirsize files from supplied patch (modified a bit).
1982 30. The use of :fail: followed by an empty string could lead Exim to respond to
1983 sender verification failures with (e.g.):
1985 550 Verification failed for <xxx>
1986 550 Sender verify failed
1988 where the first response line was missing the '-' that indicates it is not
1989 the final line of the response.
1991 31. The loop for finding the name of the user that called Exim had a hardwired
1992 limit of 10; it now uses the value of finduser_retries, which is used for
1993 all other user lookups.
1995 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1998 33. Exim was neglecting to zero errno before one call of strtol() when
1999 expanding a string and expecting an integer value. On some systems this
2000 resulted in spurious "integer overflow" errors. Also, it was casting the
2001 result into an int without checking.
2003 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2004 did not work. The code looks as if it has *never* worked, though it appears
2005 to have been documented since at least releast 1.62. I have made it work.
2007 35. The "timeout_DNS" error in retry rules, also documented since at least
2008 1.62, also never worked. As it isn't clear exactly what this means, and
2009 clearly it isn't a major issue, I have abolished the feature by treating it
2010 as "timeout", and writing a warning to the main and panic logs.
2012 36. The display of retry rules for -brt wasn't always showing the error code
2015 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2016 timeout_connect_A, timeout_connect_MX.
2018 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2019 to the empty sender.
2021 39. The daemon was not analysing the content of -oX till after it had closed
2022 stderr and disconnected from the controlling terminal. This meant that any
2023 syntax errors were only noted on the panic log, and the return code from
2024 the command was 0. By re-arranging the code a little, I've made the
2025 decoding happen first, so such errors now appear on stderr, and the return
2026 code is 1. However, the actual setting up of the sockets still happens in
2027 the disconnected process, so errors there are still only recorded on the
2030 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2031 connections (as happens on some IP stacks) was logged at start up time as
2032 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2033 it from "IPv6 and IPv4", which means that two separate sockets are being
2036 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2037 says whether AF_INET or AF_INET6 was passed as an argument.
2039 42. Exiwhat output was messed up when time zones were included in log
2042 43. Exiwhat now gives more information about the daemon's listening ports,
2043 and whether -tls-on-connect was used.
2045 44. The "port" option of the smtp transport is now expanded.
2047 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2048 silently ignored. Now an error message is written to the main and panic
2051 46. There's a new ACL modifier called "logwrite" which writes to a log file
2052 as soon as it is encountered.
2054 47. Added $local_user_uid and $local_user_gid at routing time.
2056 48. Exim crashed when trying to verify a sender address that was being
2059 49. Exim was recognizing only a space character after ".include". It now also
2060 recognizes a tab character.
2062 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2063 extracting the relevant information from the specification. The man page no
2064 longer contains scrambled data for the -d option, and I've added a section
2065 at the front about calling Exim under different names.
2067 51. Added "extra_headers" argument to the "mail" command in filter files.
2069 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2072 53. Installed eximstats 1.29.
2074 54. Added transport_filter_timeout as a generic transport option.
2076 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2077 Cc: header lines. This was required by RFC 822, but it not required by RFC
2080 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2081 incoming messages that did not have them. Now it does so only if the
2082 message originates locally, that is, if there is no associated remote host
2083 address. When Resent- header lines are present, this applies to the Resent-
2084 lines rather than the non-Resent- lines.
2086 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2087 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2089 58. Messages for configuration errors now include the name of the main
2090 configuration file - useful now that there may be more than one file in a
2091 list (.included file names were always shown).
2093 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2094 for those rare installations that do not start the daemon as root or run it
2095 setuid root. I've cut out the call to initgroups() if the daemon is not
2098 60. The Exim user and group can now be bound into the binary as text strings
2099 that are looked up at the start of Exim's processing.
2101 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2103 62. Added $mailstore_basename variable.
2105 63. Installed patch to sieve.c from Michael Haardt.
2107 64. When Exim failed to open the panic log after failing to open the main log,
2108 the original message it was trying to log was written to stderr and debug
2109 output, but if they were not available (the usual case in production), it
2110 was lost. Now it is written to syslog before the two lines that record the
2111 failures to open the logs.
2113 65. Users' Exim filters run in subprocesses under the user's uid. It is
2114 possible for a "deliver" command or an alias in a "personal" command to
2115 provoke an address rewrite. If logging of address rewriting is configured,
2116 this fails because the process is not running as root or exim. There may be
2117 a better way of dealing with this, but for the moment (because 4.30 needs
2118 to be released), I have disabled address rewrite logging when running a
2119 filter in a non-root, non-exim process.
2125 1. The buildconfig auxiliary program wasn't quoting the value set for
2126 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2127 not defined. This bug was masked in 4.22 by the effect that was fixed in
2130 2. Some messages that were rejected after a message id was allocated were
2131 shown as "incomplete" by exigrep. It no longer does this for messages that
2132 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2134 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2135 have allow_domain_literals set, the ID did not get logged in the <= line.
2136 Domain literals are now always recognized in Message-ID: header lines.
2138 4. The first argument for a ${extract expansion item is the key name or field
2139 number. Leading and trailing spaces in this item were not being ignored,
2140 causing some misleading effects.
2142 5. When deliver_drop_privilege was set, single queue runner processes started
2143 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2144 same command in the process it spins off) were not dropping privilege.
2146 6. When the daemon running as "exim" started a queue runner, it always
2147 re-executed Exim in the spun-off process. This is a waste of effort when
2148 deliver_drop_privilege is set. The new process now just calls the
2149 queue-runner function directly.
2155 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2158 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2159 ignored. Though the use of -oP was forcing the writing of a pid file, it
2160 was always written to the default place.
2162 3. If the message "no IP address found for host xxxx" is generated during
2163 incoming verification, it is now followed by identification of the incoming
2164 connection (so you can more easily find what provoked it).
2166 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2168 5. Added some features to "harden" Exim a bit more against certain attacks:
2170 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2171 be put in Local/Makefile. This is like the never_users runtime option,
2172 but it cannot be overridden. The default setting is "root".
2174 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2175 prefix string with which any file named in a -C command line option
2178 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2179 is retained for -C and -D only if the caller of Exim is root. Without
2180 it, the exim user may also use -C and -D and retain privilege.
2182 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2183 command line option is disabled.
2185 6. Macro names set by the -D option must start with an upper case letter, just
2186 like macro names defined in the configuration file.
2188 7. Added "dereference=" facility to LDAP.
2190 8. Two instances of the typo "uknown" in the source files are fixed.
2192 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2193 the Configure-Makefile script screwed up while processing it.
2195 10. Incorporated PCRE 4.4.
2197 11. The SMTP synchronization check was not operating right at the start of an
2198 SMTP session. For example, it could not catch a HELO sent before the client
2199 waited for the greeting. There is now a check for outstanding input at the
2200 point when the greeting is written. Because of the duplex, asynchronous
2201 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2202 way, but not yet received, when the check is performed.
2204 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2205 on TCP/IP sockets, because this apparently causes some broken clients to
2208 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2209 unchanged) from the Cygwin maintainer.
2211 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2212 was supported instead of testing for mbx. Effectively a typo.
2214 15. The spa authenticator server code was not checking that the input it
2215 received was valid base64.
2217 16. The debug output line for the "set" modifier in ACLs was not showing the
2218 name of the variable that was being set.
2220 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2222 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2223 lookup on its own. Something else was needed to trigger the lookup. For
2224 example, a match in host_lookup or the need for a host name in a host list.
2225 Now, if $sender_host_name is referenced and the host name has not yet been
2226 looked up, a lookup is performed. If the lookup fails, the variable remains
2227 empty, and $host_lookup_failed is set to "1".
2229 19. Added "eqi" as a case-independent comparison operator.
2231 20. The saslauthd authentication condition could segfault if neither service
2232 nor realm was specified.
2234 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2235 error message that was logged was misleading, and incoming SMTP
2236 connections were dropped. The message is now more accurate, and temporary
2237 errors are given to SMTP connections.
2239 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2240 (except RSET) with a 421 error, until QUIT is received. However, it was
2241 failing to send a response to QUIT.
2243 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2244 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2245 ACL is now run after the helo_try_verify_hosts code.
2247 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2248 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2249 case-independent, so other case variants are also recognized). Apparently
2250 some systems use these upper case variants.
2252 25. If more than two messages were waiting for the same host, and a transport
2253 filter was specified for the transport, Exim sent two messages over the
2254 same TCP/IP connection, and then failed with "socket operation on non-
2255 socket" when it tried to send the third.
2257 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2259 27. The extern definition of crypt16() in expand.c was not being excluded when
2260 the OS had its own crypt16() function.
2262 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2263 as a preferred synonym for return_size_limit, both as an option and as an
2266 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2268 30. Changed the default configuration ACL to relax the local part checking rule
2269 for addresses that are not in any local domains. For these addresses,
2270 slashes and pipe symbols are allowed within local parts, but the sequence
2271 /../ is explicitly forbidden.
2273 31. SPA server authentication was not clearing the challenge buffer before
2276 32. log_message in a "warn" ACL statement was writing to the reject log as
2277 well as to the main log, which contradicts the documentation and doesn't
2278 seem right (because no rejection is happening). So I have stopped it.
2280 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2281 However, I am unable to do any testing of this.
2283 34. Fixed an infelicity in the appendfile transport. When checking directories
2284 for a mailbox, to see if any needed to be created, it was accidentally
2285 using path names with one or more superfluous leading slashes; tracing
2286 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2288 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2289 contents are added to the log line that is written for every discarded
2290 recipient. (Previously a log_message setting was ignored.)
2292 36. The ${quote: operator now quotes the string if it is empty.
2294 37. The install script runs exim in order to find its version number. If for
2295 some reason other than non-existence or emptiness, which it checks, it
2296 could not run './exim', it was installing it with an empty version number,
2297 i.e. as "exim-". This error state is now caught, and the installation is
2300 38. An argument was missing from the function that creates an error message
2301 when Exim fails to connect to the socket for saslauthd authentication.
2302 This could cause Exim to crash, or give a corrupted message.
2304 39. Added isip, isip4, and isip6 to ${if conditions.
2306 40. The ACL variables $acl_xx are now saved with the message, and can be
2307 accessed later in routers, transports, and filters.
2309 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2310 strings in the file are not string-expanded.
2312 42. If a MAIL command specified a SIZE value that was too large to fit into an
2313 int variable, the check against message_size_limit failed. Such values are
2314 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2315 day this will have to be increased, but I don't think I want to be around
2316 when emails are that large.
2323 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2324 iconv() is not standard in FreeBSD.
2326 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2327 IPv6 enabled. The observed symptom was a segmentation fault on return from
2328 the function os_common_find_running_interfaces() in src/os.c.
2330 3. In the check_special_case() function in daemon.c I had used "errno" as an
2331 argument name, which causes warnings on some systems. This was basically a
2332 typo, since it was named "eno" in the comments!
2334 4. The code that waits for the clock to tick (at a resolution of some fraction
2335 of a second) so as to ensure message-id uniqueness was always waiting for
2336 at least one whole tick, when it could have waited for less. [This is
2337 almost certainly not relevant at current processor speeds, where it is
2338 unlikely to ever wait at all. But we try to future-proof.]
2340 5. The function that sleeps for a time interval that includes fractions of a
2341 second contained a race. It did not block SIGALRM between setting the
2342 timer, and suspending (a couple of lines later). If the interval was short
2343 and the sigsuspend() was delayed until after it had expired, the suspension
2344 never ended. On busy systems this could lead to processes getting stuck for
2347 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2348 process, before it forks any delivery processes. The open lookup caching
2349 mechanism meant that the open file or database connection was passed into
2350 the delivery process. The problem was that delivery processes always tidy
2351 up cached lookup data. This could cause a problem for the next delivery
2352 process started by the queue runner, because the external queue runner
2353 process does not know about the closure. So the next delivery process
2354 still has data in the lookup cache. In the case of a file lookup, there was
2355 no problem because closing a file descriptor in a subprocess doesn't affect
2356 the parent. However, if the lookup was caching a connection to a database,
2357 the connection was closed, and the second delivery process was likely to
2358 see errors such as "PGSQL: query failed: server closed the connection
2359 unexpectedly". The problem has been fixed by closing all cached lookups
2360 in a queue runner before running a delivery process.
2362 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2363 seem to have the "const" qualifier which it has on other OS. I've
2366 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2367 *of the same type* (client or server) with the same public name that an
2368 error should be diagnosed.
2370 9. When Exim looked up a host name for an IP address, but failed to find the
2371 original IP address when looking up the host name (a safety check), it
2372 output the message "<ip address> does not match any IP for NULL", which was
2373 confusing, to say the least. The bug was that the host name should have
2374 appeared instead of "NULL".
2376 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2377 user that is built into the binary, and the -C or -D options is used, root
2378 privilege is dropped before the configuration file is read. In addition,
2379 logging is switched to stderr instead of the normal log files. If the
2380 configuration then re-defines the Exim user, the unprivileged environment
2381 is probably not what is expected, so Exim logs a panic warning message (but
2384 However, if deliver_drop_privilege is set, the unprivileged state may well
2385 be exactly what is intended, so the warning has been cut out in that case,
2386 and Exim is allowed to try to write to its normal log files.
2392 1. smtp_return_error_details was not giving details for temporary sender
2393 or receiver verification errors.
2395 2. Diagnose a configuration error if two authenticators have the same public
2398 3. Exim used not to create the message log file for a message until the first
2399 delivery attempt. This could be confusing when incoming messages were held
2400 for policy or load reasons. The message log file is now created at the time
2401 the message is received, and an initial "Received" line is written to it.
2403 4. The automatically generated man page for command line options had a minor
2404 bug that caused no ill effects; however, a more serious problem was that
2405 the procedure for building the man page automatically didn't always
2406 operate. Consequently, release 4.20 contains an out-of-date version. This
2407 shouldn't happen again.
2409 5. When building Exim with embedded Perl support, the script that builds the
2410 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2411 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2413 6. The freeze_tell option was not being used for messages that were frozen on
2414 arrival, either by an ACL or by local_scan().
2416 7. Added the smtp_incomplete_transaction log selector.
2418 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2419 was accepting AUTH without a new EHLO.
2421 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2422 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2425 10. Logging of TCP/IP connections (when configured) now happens in the main
2426 daemon process instead of the child process, so that the TCP/IP connection
2427 count is more accurate (but it can never be perfect).
2429 11. The use of "drop" in a nested ACL was not being handled correctly in the
2430 outer ACL. Now, if condition failure induced by the nested "drop" causes
2431 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2432 or "require"), the connection is dropped.
2434 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2435 that yield "discard" can now be used with an "accept" or a "discard" verb,
2436 but an error is generated for any others (because I can't see a useful way
2437 to define what should happen).
2439 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2440 are now processed in the same way as lines in the Exim configuration file.
2441 In particular, continuation lines are supported.
2443 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2445 15. Added -ti meaning -t -i.
2447 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2448 domains, and warn by logging if others are found.
2450 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2451 returned from GIFCONF when Exim is trying to find the list of interfaces on
2452 a host. The code in os.c has been modified to copy each ifreq to an aligned
2453 structure in all cases.
2455 Also, in some cases, the returned ifreq's were being copied to a 'struct
2456 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2457 means the last couple of bytes of an IPv6 address could be chopped if the
2458 ifreq contained only a normal sockaddr (14 bytes storage).
2460 18. Named domain lists were not supported in the hosts_treat_as_local option.
2461 An entry such as +xxxx was not recognized, and was treated as a literal
2464 19. Ensure that header lines added by a DATA ACL are included in the reject log
2465 if the ACL subsequently rejects the message.
2467 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2468 MD5 (which is deprecated).
2470 21. When testing a filter file using -bf, Exim was writing a message when it
2471 took the sender from a "From " line in the message, but it was not doing so
2472 when it took $return_path from a Return-Path: header line. It now does.
2474 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2475 with a valid header line field name (a series of printing characters
2476 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2478 23. Changed "disc" in the source to "disk" to conform to the documentation and
2479 the book and for uniformity.
2481 24. Ignore Sendmail's -Ooption=value command line item.
2483 25. When execve() failed while trying to run a command in a pipe transport,
2484 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2485 could be confused with a return value of 69 from the command itself. This
2486 has been changed to 127, the value the shell returns if it is asked to run
2487 a non-existent command. The wording for the related log line suggests a
2488 non-existent command as the problem.
2490 26. If received_header_text expands to an empty string, do not add a Received:
2491 header line to the message. (Well, it adds a token one on the spool, but
2492 marks it "old" so that it doesn't get used or transmitted.)
2494 27. Installed eximstats 1.28 (addition of -nt option).
2496 28. There was no check for failure on the call to getsockname() in the daemon
2497 code. This can fail if there is a shortage of resources on the system, with
2498 ENOMEM, for example. A temporary error is now given on failure.
2500 29. Contrary to the C standard, it seems that in some environments, the
2501 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2502 program. Exim now does this explicitly; it affects the formatting of
2503 timestamps using strftime().
2505 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2506 complaints. I've now initialized all the variables, to avoid this.
2508 32. Header lines added by a system filter were not being "seen" during
2509 transport-time rewrites.
2511 33. The info_callback() function passed to OpenSSL is set up with type void
2512 (*)(SSL *, int, int), as described somewhere. However, when calling the
2513 function (actually a macro) that sets it up, the type void(*)() is
2514 expected. I've put in a cast to prevent warnings from picky compilers.
2516 34. If a DNS black list lookup found a CNAME record, but there were no A
2517 records associated with the domain it pointed at, Exim crashed.
2519 35. If a DNS black list lookup returned more than one A record, Exim ignored
2520 all but the first. It now scans all returned addresses if a particular IP
2521 value is being sought. In this situation, the contents of the
2522 $dnslist_value variable are a list of all the addresses, separated by a
2525 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2526 to accept a host name and all its aliases if the forward lookup for any of
2527 them yielded the IP address of the incoming connection. Now it accepts only
2528 those names whose forward lookup yields the correct IP address. Any other
2529 names are discarded. This closes a loophole whereby a rogue DNS
2530 administrator could create reverse DNS records to break through a
2531 wildcarded host restriction in an ACL.
2533 37. If a user filter or a system filter that ran in a subprocess used any of
2534 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2535 the wrong values were passed to the pipe command ($thisaddress had the
2536 value of $0, $0 had the value of $1, etc). This bug was introduced by
2537 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2539 38. Improved the line breaking for long SMTP error messages from ACLs.
2540 Previously, if there was no break point between 40 and 75 characters, Exim
2541 left the rest of the message alone. Two changes have been made: (a) I've
2542 reduced the minimum length to 35 characters; (b) if it can't find a break
2543 point between 35 and 75 characters, it looks ahead and uses the first one
2544 that it finds. This may give the occasional overlong line, but at least the
2545 remaining text gets split now.
2547 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2548 file descriptors might be low, and that setting 1000 would always raise it.
2549 It turns out that in some environments, the limit is already over 1000 and
2550 that lowering it causes trouble. So now Exim takes care not to decrease it.
2552 40. When delivering a message, the value of $return_path is set to $sender_
2553 address at the start of routing (routers may change the value). By an
2554 oversight, this default was not being set up when an address was tested by
2555 -bt or -bv, which affected the outcome if any router or filter referred to
2558 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2559 log only when "message" or "log_message" are set. However, if one of the
2560 conditions was an address verification, or a call to a nested ACL, the
2561 messages generated by the underlying test were being passed through. This
2562 no longer happens. The underlying message is available in $acl_verify_
2563 message for both "message" and "log_message" expansions, so it can be
2564 passed through if needed.
2566 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2567 new expansion $bh_ to give the encoded byte string without charset
2568 translation. Translation happens only if iconv() is available; HAVE_ICONV
2569 indicates this at build time. HEADERS_CHARSET gives the charset to
2570 translate to; headers_charset can change it in the configuration, and
2571 "headers charset" can change it in an individual filter file.
2573 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2574 that creates RFC 2047 encoded "words" labels them as that charset instead
2575 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2576 expansion operator; (ii) when Exim creates a From: line for a local
2577 message; (iii) when a header line is rewritten to include a "phrase" part.
2579 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2580 buggy, causing it to skip the first lines of messages whose message ID
2581 ended in 'D'. This would not have bitten before Exim release 4.14, because
2582 message IDs were unlikely to end in 'D' before then. The effect was to have
2583 incorrect size information for certain domains.
2585 45. #include "config.h" was missing at the start of the crypt16.c module. This
2586 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2589 46. If there was a timeout during a "random" callout check, Exim treated it as
2590 a failure of the random address, and carried on sending RSET and the real
2591 address. If the delay was just some slowness somewhere, the response to the
2592 original RCPT would be taken as a response to RSET and so on, causing
2593 mayhem of various kinds.
2595 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2596 when I implemented it. It didn't allow for the fact that some option values
2597 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2598 the right test for positive values.
2600 48. Domain names in DNS records are case-independent. Exim always looks them up
2601 in lower case. Some resolvers return domain names in exactly the case they
2602 appear in the zone file, that is, they may contain uppercase letters. Not
2603 all resolvers do this - some return always lower case. Exim was treating a
2604 change of case by a resolver as a change of domain, similar to a widening
2605 of a domain abbreviation. This triggered its re-routing code and so it was
2606 trying to route what was effectively the same domain again. This normally
2607 caused routing to fail (because the router wouldn't handle the domain
2608 twice). Now Exim checks for this case specially, and just changes the
2609 casing of the domain that it ultimately uses when it transmits the message
2612 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2615 50. If a filter generated a file delivery with a non-absolute name (possible if
2616 no home directory exists for the router), the forbid_file option was not
2619 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2620 the existing equality matching.
2622 52. Exim was using ints instead of ino_t variables in some places where it was
2623 dealing with inode numbers.
2625 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2626 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2627 called TMPDIR, and if it finds it is different, it changes its value.
2629 54. The smtp_printf() function is now made available to local_scan() so
2630 additional output lines can be written before returning. There is also an
2631 smtp_fflush() function to enable the detection of a dropped connection.
2632 The variables smtp_input and smtp_batched_input are exported to
2635 55. Changed the default runtime configuration: the message "Unknown user"
2636 has been removed from the ACL, and instead placed on the localuser router,
2637 using the cannot_route_message feature. This means that any verification
2638 failures that generate their own messages won't get overridden. Similarly,
2639 the "Unrouteable address" message that was in the ACL for unverifiable
2640 relay addresses has also been removed.
2642 56. Added hosts_avoid_esmtp to the smtp transport.
2644 57. The exicyclog script was not checking for the esoteric option
2645 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2646 will work only if exicyclog is run under the appropriate euid.
2648 58. Following a discussion on the list, the rules by which Exim recognises line
2649 endings on incoming messages have been changed. The -dropcr and drop_cr
2650 options are now no-ops, retained only for backwards compatibility. The
2651 following line terminators are recognized: LF CRLF CR. However, special
2652 processing applies to CR:
2654 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2655 nor a local message in the state where . is a terminator.
2657 (ii) If a bare CR is encountered in a header line, an extra space is added
2658 after the line terminator so as not to end the header. The reasoning
2659 behind this is that bare CRs in header lines are most likely either
2660 to be mistakes, or people trying to play silly games.
2662 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2663 was being incorrectly given as 18 bytes larger than it should have been.
2664 This is a VOB (very old bug).
2666 60. This may never have affected anything current, but just in case it has:
2667 When the local host is found other than at the start of a list of hosts,
2668 the local host, those with the same MX, and any that follow, are discarded.
2669 When the list in question was part of a longer list of hosts, the following
2670 hosts (not currently being processed) were also being discarded. This no
2671 longer happens. I'm not sure if this situation could ever has previously
2674 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2677 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2678 header. This is something that is recommended in a new Internet Draft, and
2679 is something that is documented as being done by Sendmail. There are two
2680 possible values. For messages generated by the autoreply transport, Exim
2683 Auto-Submitted: auto-replied
2685 whereas for all other generated messages (e.g. bounces) it adds
2687 Auto-Submitted: auto-generated
2689 63. The "personal" condition in filters now includes a test for the
2690 Auto-Submitted: header. If it contains the string "auto-" the message it
2691 not considered personal.
2693 64. Added rcpt_include_affixes as a generic transport option.
2695 65. Added queue_only_override (default true).
2697 66. Added the syslog_duplication option.
2699 67. If what should have been the first header line of a message consisted of
2700 a space followed by a colon, Exim was mis-interpreting it as a header line.
2701 It isn't of course - it is syntactically invalid and should therefore be
2702 treated as the start of the message body. The misbehaviour could have
2703 caused a number of strange effects, including loss of data in subsequent
2704 header lines, and spool format errors.
2706 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2707 client host had authenticated. This control can now be exercised by an ACL
2708 for more flexibility.
2710 69. By default, callouts do not happen when testing with -bh. There is now a
2711 variant, -bhc, which does actually run the callout code, including
2712 consulting and updating the callout cache.
2714 70. Added support for saslauthd authentication, courtesy of Alexander
2717 71. If statvfs() failed on the spool or log directories while checking their
2718 size for availability, Exim confusingly gave the error "space shortage".
2719 Furthermore, in debugging mode it crashed with a floating point exception.
2720 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2721 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2722 serious problem, Exim now writes to the main and panic logs when this
2723 happens, with details of the failure. It then refuses to accept the
2724 incoming message, giving the message "spool directory problem" or "log
2725 directory problem" with a 421 code for SMTP messages.
2727 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2728 0, 1, and 2 exist, because some OS complain for execs without them (see
2729 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2730 descriptors. However, the code omitted to check that the open succeeded,
2731 causing mysterious errors if for some reason the permissions on /dev/null
2732 got screwed. Now Exim writes a message to the main and panic logs, and
2733 bombs out if it can't open /dev/null.
2735 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2736 interact so that it is all more flexible. It is supposed to remain
2737 backwards compatible. Also added extra_local_interfaces.
2739 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2740 to bomb out with an assertion failure - to the client this appears as a
2741 connection drop. This problem occurs in the part of the code that was taken
2742 from the Samba project. Fortunately, the assertion is in a very simple
2743 function, so I have fixed this by reproducing the function inline in the
2744 one place where it is called, and arranging for authentication to fail
2745 instead of killing the process with assert().
2747 75. The SPA client code was not working when the server requested OEM rather
2748 than Unicode encoding.
2750 76. Added code to make require_files with a specific uid setting more usable in
2751 the case where statting the file as root fails - usually a non-root-mounted
2752 NFS file system. When this happens and the failure is EACCES, Exim now
2753 forks a subprocess and does the per-uid checking as the relevant uid.
2755 77. Added process_log_path.
2757 78. If log_file_path was not explicitly set, a setting of check_log_space or
2758 check_log_inodes was ignored.
2760 79. If a space check for the spool or log partitions fails, the incident is now
2761 logged. Of course, in the latter case the data may get lost...
2763 80. Added the %p formatting code to string_format() so that it can be used to
2764 print addresses in debug_print(). Adjusted all the address printing in the
2765 debugging in store.c to use %p rather than %d.
2767 81. There was a concern that a line of code in smtp_in.c could overflow a
2768 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2769 initially expressed, the concern was not well-founded, because trailing
2770 spaces are removed early. However, if the trailing spaces were followed by
2771 a NULL, they did not get removed, so the overflow was possible. Two fixes
2774 (a) I re-wrote the offending code in a cleaner fashion.
2775 (b) If an incoming SMTP command contains a NULL character, it is rejected
2778 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2779 runs initgroups(), so that if the Exim user is in any additional groups,
2780 they will be used during message reception.
2786 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob