1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.119 2005/04/06 16:43:59 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
161 PH/27 $acl_verify_message is now set immediately after the failure of a
162 verification in an ACL, and so is available in subsequent modifiers. In
163 particular, the message can be preserved by coding like this:
165 warn !verify = sender
166 set acl_m0 = $acl_verify_message
168 Previously, $acl_verify_message was set only while expanding "message"
169 and "log_message" when a very denied access.
171 PH/28 Modified OS/os.c-Linux with
173 -#ifndef OS_LOAD_AVERAGE
174 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
176 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
177 nomenclature these days.)
179 PH/29 Installed patch from the Sieve maintainer that adds the options
180 sieve_useraddress and sieve_subaddress to the redirect router.
182 PH/30 In these circumstances:
183 . Two addresses routed to the same list of hosts;
184 . First host does not offer TLS;
185 . First host accepts first address;
186 . First host gives temporary error to second address;
187 . Second host offers TLS and a TLS session is established;
188 . Second host accepts second address.
189 Exim incorrectly logged both deliveries with the TLS parameters (cipher
190 and peerdn, if requested) that were in fact used only for the second
193 PH/31 When doing a callout as part of verifying an address, Exim was not paying
194 attention to any local part prefix or suffix that was matched by the
195 router that accepted the address. It now behaves in the same way as it
196 does for delivery: the affixes are removed from the local part unless
197 rcpt_include_affixes is set on the transport.
199 PH/32 Add the sender address, as F=<...>, to the log line when logging a
200 timeout during the DATA phase of an incoming message.
204 A note about Exim versions 4.44 and 4.50
205 ----------------------------------------
207 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
208 changes of various kinds. As a consequence, a big documentation update was
209 needed. This delayed the release for rather longer than seemed good, especially
210 in the light of a couple of (minor) security issues. Therefore, the changes
211 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
212 release. So 4.44 and 4.50 are in effect two different branches that both start
215 I have left the 4.50 change log unchanged; it contains all the changes since
216 4.43. The change log for 4.44 is below; many of its items are identical to
217 those for 4.50. This seems to be the most sensible way to preserve the
218 historical information.
224 1. Minor wording change to the doc/README.SIEVE file.
226 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
227 computation of the current number of files was incorrect.
229 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
230 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
233 4. Give more explanation in the error message when the command for a transport
234 filter fails to execute.
236 5. There are several places where Exim runs a non-Exim command in a
237 subprocess. The SIGUSR1 signal should be disabled for these processes. This
238 was being done only for the command run by the queryprogram router. It is
239 now done for all such subprocesses. The other cases are: ${run, transport
240 filters, and the commands run by the lmtp and pipe transports.
242 6. Added CONFIGURE_GROUP build-time option.
244 7. Some older OS have a limit of 256 on the maximum number of file
245 descriptors. Exim was using setrlimit() to set 1000 as a large value
246 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
247 systems. I've change it so that if it can't get 1000, it tries for 256.
249 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
250 was an oversight, and furthermore, ever since the addition of extra
251 controls (e.g. 4.43/32), the checks on when to allow different forms of
252 "control" were broken. There should now be diagnostics for all cases when a
253 control that does not make sense is encountered.
255 9. Added the /retain_sender option to "control=submission".
257 10. $recipients is now available in the predata ACL (oversight).
259 11. Tidy the search cache before the fork to do a delivery from a message
260 received from the command line. Otherwise the child will trigger a lookup
261 failure and thereby defer the delivery if it tries to use (for example) a
262 cached ldap connection that the parent has called unbind on.
264 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
265 of $address_data from the recipient verification was clobbered by the
268 13. The value of address_data from a sender verification is now available in
269 $sender_address_data in subsequent conditions in the ACL statement.
271 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
273 15. Added a new option "connect=<time>" to callout options, to set a different
276 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
277 was its contents. (It was OK if the option was not defined at all.)
279 17. A "Completed" log line is now written for messages that are removed from
280 the spool by the -Mrm option.
282 18. New variables $sender_verify_failure and $recipient_verify_failure contain
283 information about exactly what failed.
285 19. Added -dd to debug only the daemon process.
287 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
288 handles timeouts, both on the server side and network timeouts. Renamed the
289 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
291 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
292 It is now set to "smtps".
294 22. $host_address is now set to the target address during the checking of
297 23. When checking ignore_target_hosts for an ipliteral router, no host name was
298 being passed; this would have caused $sender_host_name to have been used if
299 matching the list had actually called for a host name (not very likely,
300 since this list is usually IP addresses). A host name is now passed as
303 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
304 code that specifies a non-restarting handler (typically sigaction() in
305 modern systems) in an attempt to fix a rare and obscure crash bug.
307 25. Narrowed the window for a race in the daemon that could cause it to ignore
308 SIGCHLD signals. This is not a major problem, because they are used only to
309 wake it up if nothing else does.
311 26. A malformed maildirsize file could cause Exim to calculate negative values
312 for the mailbox size or file count. Odd effects could occur as a result.
313 The maildirsize information is now recalculated if the size or filecount
316 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
317 support for a long time. Removed HAVE_SYS_VFS_H.
319 28. Installed the latest version of exipick from John Jetmore.
321 29. In an address list, if the pattern was not a regular expression, an empty
322 subject address (from a bounce message) matched only if the pattern was an
323 empty string. Non-empty patterns were not even tested. This was the wrong
324 because it is perfectly reasonable to use an empty address as part of a
325 database query. An empty address is now tested by patterns that are
326 lookups. However, all the other forms of pattern expect the subject to
327 contain a local part and a domain, and therefore, for them, an empty
328 address still always fails if the pattern is not itself empty.
330 30. Exim went into a mad DNS loop when attempting to do a callout where the
331 host was specified on an smtp transport, and looking it up yielded more
334 31. Re-factored the code for checking spool and log partition space into a
335 function that finds that data and another that does the check. The former
336 is then used to implement four new variables: $spool_space, $log_space,
337 $spool_inodes, and $log_inodes.
339 32. The RFC2047 encoding function was originally intended for short strings
340 such as real names; it was not keeping to the 75-character limit for
341 encoded words that the RFC imposes. It now respects the limit, and
342 generates multiple encoded words if necessary. To be on the safe side, I
343 have increased the buffer size for the ${rfc2047: expansion operator from
346 33. It is now permitted to omit both strings after an "if" condition; if the
347 condition is true, the result is "true". As before, when the second string
348 is omitted, a false condition yields an empty string. This makes it less
349 cumbersome to write custom ACL and router conditions.
351 34. Failure to deliver a bounce message always caused it to be frozen, even if
352 there was an errors_to setting on the router. The errors_to setting is now
355 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
356 canonical form (fully expanded) before being placed in
357 $sender_host_address.
359 36. The table in the code that translates DNS record types into text (T_A to
360 "A" for instance) was missing entries for NS and CNAME. It is just possible
361 that this could have caused confusion if both these types were looked up
362 for the same domain, because the text type is used as part of Exim's
363 per-process caching. But the chance of anyone hitting this buglet seems
366 37. The dnsdb lookup has been extended in a number of ways.
368 (1) There is a new type, "zns", which walks up the domain tree until it
369 finds some nameserver records. It should be used with care.
371 (2) There is a new type, "mxh", which is like "mx" except that it returns
372 just the host names, not the priorities.
374 (3) It is now possible to give a list of domains (or IP addresses) to be
375 looked up. The behaviour when one of the lookups defers can be
376 controlled by a keyword.
378 (4) It is now possible to specify the separator character for use when
379 multiple records are returned.
381 38. The dnslists ACL condition has been extended: it is now possible to supply
382 a list of IP addresses and/or domains to be looked up in a particular DNS
385 39. Added log_selector=+queue_time_overall.
387 40. When running the queue in the test harness, wait just a tad after forking a
388 delivery process, to get repeatability of debugging output.
390 41. Include certificate and key file names in error message when GnuTLS fails
391 to set them up, because the GnuTLS error message doesn't include the name
392 of the failing file when there is a problem reading it.
394 42. Allow both -bf and -bF in the same test run.
396 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
398 44. The "Exiscan patch" is now merged into the mainline Exim source.
400 45. Sometimes the final signoff response after QUIT could fail to get
401 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
402 < 0 before doing a fflush(). This bug looks as though it goes back to the
403 introduction of TLS in release 3.20, but "sometimes" must have been rare
404 because the tests only now provoked it.
406 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
407 (this can affect the format of dates).
409 47. exim_tidydb, when checking for the continued existence of a message for
410 which it has found a message-specific retry record, was not finding
411 messages that were in split spool directories. Consequently, it was
412 deleting retry records that should have stayed in existence.
414 48. Steve fixed some bugs in eximstats.
416 49. The SPA authentication driver was not abandoning authentication and moving
417 on to the next authenticator when an expansion was forced to fail,
418 contradicting the general specification for all authenticators. Instead it
419 was generating a temporary error. It now behaves as specified.
421 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
422 (the order specifies the preference for clients). The order is now AES256,
423 AES128, 3DES, ARCFOUR128.
425 51. Small patch to Sieve code - explicitly set From: when generating an
428 52. Exim crashed if a remote delivery caused a very long error message to be
429 recorded - for instance if somebody sent an entire SpamAssassin report back
430 as a large number of 550 error lines. This bug was coincidentally fixed by
431 increasing the size of one of Exim's internal buffers (big_buffer) that
432 happened as part of the Exiscan merge. However, to be on the safe side, I
433 have made the code more robust (and fixed the comments that describe what
436 53. Now that there can be additional text after "Completed" in log lines (if
437 the queue_time_overall log selector is set), a one-byte patch to exigrep
438 was needed to allow it to recognize "Completed" as not the last thing in
441 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
442 patch that reportedly fixes this has been added. I am not expert enough to
443 create a test for it. This is what the patch creator wrote:
445 "I found a little strange behaviour of ldap code when working with
446 Windows 2003 AD Domain, where users was placed in more than one
447 Organization Units. When I tried to give exim partial DN, the exit code
448 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
449 But simultaneously result of request was absolutely normal ldap result,
450 so I produce this patch..."
452 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
453 REFERENCE, so I have modified the code to exclude the patch when that macro
456 55. Some experimental protocols are using DNS PTR records for new purposes. The
457 keys for these records are domain names, not reversed IP addresses. The
458 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
459 leaves it alone. Component reversal etc. now happens only for IP addresses.
462 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
464 57. Double the size of the debug message buffer (to 2048) so that more of very
465 long debug lines gets shown.
467 58. The exicyclog utility now does better if the number of log files to keep
468 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
470 59. Two changes related to the smtp_active_hostname option:
472 (1) $smtp_active_hostname is now available as a variable.
473 (2) The default for smtp_banner uses $smtp_active_hostname instead
474 of $primary_hostname.
476 60. The host_aton() function is supposed to be passed a string that is known
477 to be a valid IP address. However, in the case of IPv6 addresses, it was
478 not checking this. This is a hostage to fortune. Exim now panics and dies
479 if the condition is not met. A case was found where this could be provoked
480 from a dnsdb PTR lookup with an IPv6 address that had more than 8
481 components; fortuitously, this particular loophole had already been fixed
482 by change 4.50/55 above.
484 If there are any other similar loopholes, the new check in host_aton()
485 itself should stop them being exploited. The report I received stated that
486 data on the command line could provoke the exploit when Exim was running as
487 exim, but did not say which command line option was involved. All I could
488 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
492 61. There was a buffer overflow vulnerability in the SPA authentication code
493 (which came originally from the Samba project). I have added a test to the
494 spa_base64_to_bits() function which I hope fixes it.
497 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
498 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
500 63. The daemon start-up calls getloadavg() while still root for those OS that
501 need the first call to be done as root, but it missed one case: when
502 deliver_queue_load_max is set with deliver_drop_privilege. This is
503 necessary for the benefit of the queue runner, because there is no re-exec
504 when deliver_drop_privilege is set.
506 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
509 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
510 was in use, was not putting the data itself into the right store pool;
511 consequently, it could be overwritten for a subsequent message in the same
512 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
515 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
517 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
518 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
519 the function were treating the return as a boolean value, which happened to
520 work because 0=false and not-0=true, but is not correct code.
522 68. The host_aton() function was not handling scoped IPv6 addresses (those
523 with, for example, "%eth0" on the end) correctly.
525 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
526 negated items (that is, ~something) in unsigned ints. Some compilers
527 apparently mutter when there is no cast.
529 70. If an address verification called from an ACL failed, and did not produce a
530 user-specific message (i.e. there was only a "system" message), nothing was
531 put in $acl_verify_message. In this situation, it now puts the system
534 71. Change 4.23/11 added synchronization checking at the start of an SMTP
535 session; change 4.31/43 added the unwanted input to the log line - except
536 that it did not do this in the start of session case. It now does.
538 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
539 This is wrong and can cause the other end to generate a synchronization
540 error if it is another Exim or anything else that does the synchronization
541 check. A QUIT command is no longer sent after a timeout.
543 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
546 74. The defer_ok option of callout verification was not working if it was used
547 when verifying addresses in header lines, that is, for this case:
549 verify = header_sender/callout=defer_ok
551 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
552 those file descriptors could be used for SMTP connections. If anything
553 wrote to stderr (the example that came up was "warn" in embedded Perl), it
554 could be sent to the SMTP client, causing chaos. The daemon now opens
555 stdin, stdout, and stderr to /dev/null when it puts itself into the
558 76. Arrange for output from Perl's "warn" command to be written to Exim's main
559 log by default. The user can override this with suitable Perl magic.
561 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
562 the log message when discard triggers, was not working for the DATA ACL or
563 for the non-SMTP ACL.
565 78. Error message wording change in sieve.c.
567 79. If smtp_accept_max_per_host was set, the number of connections could be
568 restricted to fewer than expected, because the daemon was trying to set up
569 a new connection before checking whether the processes handling previous
570 connections had finished. The check for completed processes is now done
571 earlier. On busy systems, this bug wouldn't be noticed because something
572 else would have woken the daemon, and it would have reaped the completed
575 80. If a message was submitted locally by a user whose login name contained one
576 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
577 It caused a spool format error. I have fixed the spool reading code. A
578 related problem was that the "from" clause in the Received: line became
579 illegal because of the space(s). It is now covered by ${quote_local_part.
581 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
584 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
586 83. Patch from Sieve maintainer to fix unterminated string problem in
589 84. Some minor changes to the Linux configuration files to help with other
590 OS variants using glibc.
592 85. One more patch for Sieve to update vacation handling to latest spec.
595 ----------------------------------------------------
596 See the note above about the 4.44 and 4.50 releases.
597 ----------------------------------------------------
603 1. Change 4.43/35 introduced a bug that caused file counts to be
604 incorrectly computed when quota_filecount was set in an appendfile
607 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
608 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
611 3. Give more explanation in the error message when the command for a transport
612 filter fails to execute.
614 4. There are several places where Exim runs a non-Exim command in a
615 subprocess. The SIGUSR1 signal should be disabled for these processes. This
616 was being done only for the command run by the queryprogram router. It is
617 now done for all such subprocesses. The other cases are: ${run, transport
618 filters, and the commands run by the lmtp and pipe transports.
620 5. Some older OS have a limit of 256 on the maximum number of file
621 descriptors. Exim was using setrlimit() to set 1000 as a large value
622 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
623 systems. I've change it so that if it can't get 1000, it tries for 256.
625 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
626 was an oversight, and furthermore, ever since the addition of extra
627 controls (e.g. 4.43/32), the checks on when to allow different forms of
628 "control" were broken. There should now be diagnostics for all cases when a
629 control that does not make sense is encountered.
631 7. $recipients is now available in the predata ACL (oversight).
633 8. Tidy the search cache before the fork to do a delivery from a message
634 received from the command line. Otherwise the child will trigger a lookup
635 failure and thereby defer the delivery if it tries to use (for example) a
636 cached ldap connection that the parent has called unbind on.
638 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
639 of $address_data from the recipient verification was clobbered by the
642 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
643 was its contents. (It was OK if the option was not defined at all.)
645 11. A "Completed" log line is now written for messages that are removed from
646 the spool by the -Mrm option.
648 12. $host_address is now set to the target address during the checking of
651 13. When checking ignore_target_hosts for an ipliteral router, no host name was
652 being passed; this would have caused $sender_host_name to have been used if
653 matching the list had actually called for a host name (not very likely,
654 since this list is usually IP addresses). A host name is now passed as
657 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
658 code that specifies a non-restarting handler (typically sigaction() in
659 modern systems) in an attempt to fix a rare and obscure crash bug.
661 15. Narrowed the window for a race in the daemon that could cause it to ignore
662 SIGCHLD signals. This is not a major problem, because they are used only to
663 wake it up if nothing else does.
665 16. A malformed maildirsize file could cause Exim to calculate negative values
666 for the mailbox size or file count. Odd effects could occur as a result.
667 The maildirsize information is now recalculated if the size or filecount
670 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
671 support for a long time. Removed HAVE_SYS_VFS_H.
673 18. Updated exipick to current release from John Jetmore.
675 19. Allow an empty sender to be matched against a lookup in an address list.
676 Previously the only cases considered were a regular expression, or an
679 20. Exim went into a mad DNS lookup loop when doing a callout where the
680 host was specified on the transport, if the DNS lookup yielded more than
683 21. The RFC2047 encoding function was originally intended for short strings
684 such as real names; it was not keeping to the 75-character limit for
685 encoded words that the RFC imposes. It now respects the limit, and
686 generates multiple encoded words if necessary. To be on the safe side, I
687 have increased the buffer size for the ${rfc2047: expansion operator from
690 22. Failure to deliver a bounce message always caused it to be frozen, even if
691 there was an errors_to setting on the router. The errors_to setting is now
694 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
695 canonical form (fully expanded) before being placed in
696 $sender_host_address.
698 24. Updated eximstats to version 1.33
700 25. Include certificate and key file names in error message when GnuTLS fails
701 to set them up, because the GnuTLS error message doesn't include the name
702 of the failing file when there is a problem reading it.
704 26. Expand error message when OpenSSL has problems setting up cert/key files.
707 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
708 (this can affect the format of dates).
710 28. exim_tidydb, when checking for the continued existence of a message for
711 which it has found a message-specific retry record, was not finding
712 messages that were in split spool directories. Consequently, it was
713 deleting retry records that should have stayed in existence.
715 29. eximstats updated to version 1.35
716 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
717 1.35 - bugfix such that pie charts by volume are generated correctly
719 30. The SPA authentication driver was not abandoning authentication and moving
720 on to the next authenticator when an expansion was forced to fail,
721 contradicting the general specification for all authenticators. Instead it
722 was generating a temporary error. It now behaves as specified.
724 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
725 (the order specifies the preference for clients). The order is now AES256,
726 AES128, 3DES, ARCFOUR128.
728 31. Small patch to Sieve code - explicitly set From: when generating an
731 32. Exim crashed if a remote delivery caused a very long error message to be
732 recorded - for instance if somebody sent an entire SpamAssassin report back
733 as a large number of 550 error lines. This bug was coincidentally fixed by
734 increasing the size of one of Exim's internal buffers (big_buffer) that
735 happened as part of the Exiscan merge. However, to be on the safe side, I
736 have made the code more robust (and fixed the comments that describe what
739 33. Some experimental protocols are using DNS PTR records for new purposes. The
740 keys for these records are domain names, not reversed IP addresses. The
741 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
742 leaves it alone. Component reversal etc. now happens only for IP addresses.
745 34. The host_aton() function is supposed to be passed a string that is known
746 to be a valid IP address. However, in the case of IPv6 addresses, it was
747 not checking this. This is a hostage to fortune. Exim now panics and dies
748 if the condition is not met. A case was found where this could be provoked
749 from a dnsdb PTR lookup with an IPv6 address that had more than 8
750 components; fortuitously, this particular loophole had already been fixed
751 by change 4.50/55 or 4.44/33 above.
753 If there are any other similar loopholes, the new check in host_aton()
754 itself should stop them being exploited. The report I received stated that
755 data on the command line could provoke the exploit when Exim was running as
756 exim, but did not say which command line option was involved. All I could
757 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
761 35. There was a buffer overflow vulnerability in the SPA authentication code
762 (which came originally from the Samba project). I have added a test to the
763 spa_base64_to_bits() function which I hope fixes it.
766 36. The daemon start-up calls getloadavg() while still root for those OS that
767 need the first call to be done as root, but it missed one case: when
768 deliver_queue_load_max is set with deliver_drop_privilege. This is
769 necessary for the benefit of the queue runner, because there is no re-exec
770 when deliver_drop_privilege is set.
772 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
773 was in use, was not putting the data itself into the right store pool;
774 consequently, it could be overwritten for a subsequent message in the same
775 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
778 38. Sometimes the final signoff response after QUIT could fail to get
779 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
780 < 0 before doing a fflush(). This bug looks as though it goes back to the
781 introduction of TLS in release 3.20, but "sometimes" must have been rare
782 because the tests only now provoked it.
788 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
789 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
790 Now it returns an integer. A number of calls were still expecting a T/F
791 return. Fortuitously, in all cases, the tests worked in OK situations,
792 which is the norm. However, things would have gone wrong on any write
793 failures on the smtp file descriptor. This function is used when sending
794 messages over SMTP and also when doing verify callouts.
796 2. When Exim is called to do synchronous delivery of a locally submitted
797 message (the -odf or -odi options), it no longer closes stderr before doing
800 3. Implemented the mua_wrapper option.
802 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
804 5. Implemented the functions header_remove(), header_testname(),
805 header_add_at_position(), and receive_remove_recipient(), and exported them
808 6. If an ACL "warn" statement specified the addition of headers, Exim already
809 inserted X-ACL-Warn: at the start if there was no header name. However, it
810 was not making this test for the second and subsequent header lines if
811 there were newlines in the string. This meant that an invalid header could
812 be inserted if Exim was badly configured.
814 7. Allow an ACL "warn" statement to add header lines at the start or after all
815 the Received: headers, as well as at the end.
817 8. Added the rcpt_4xx retry error code.
819 9. Added postmaster_mailfrom=xxx to callout verification option.
821 10. Added mailfrom=xxxx to the callout verification option, for verify=
824 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
825 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
827 12. Inserted some casts to stop certain compilers warning when using pointer
828 differences as field lengths or precisions in printf-type calls (mostly
829 affecting debugging statements).
831 13. Added optional readline() support for -be (dynamically loaded).
833 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
834 same clock tick as a message's arrival, so that its received time was the
835 same as the "first fail" time on the retry record, and that message
836 remained on the queue past the ultimate address timeout, every queue runner
837 would try a delivery (because it was past the ultimate address timeout) but
838 after another failure, the ultimate address timeout, which should have then
839 bounced the address, did not kick in. This was a "< instead of <=" error;
840 in most cases the first failure would have been in the next clock tick
841 after the received time, and all would be well.
843 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
844 being recognized when the domain list was tested by the match_domain
845 condition in an expansion string.
847 16. Added the ${str2b64: operator.
849 17. Exim was always calling setrlimit() to set a large limit for the number of
850 processes, without checking whether the existing limit was already
851 adequate. (It did check for the limit on file descriptors.) Furthermore,
852 errors from getrlimit() and setrlimit() were being ignored. Now they are
853 logged to the main and panic logs, but Exim does carry on, to try to do its
854 job under whatever limits there are.
856 18. Imported PCRE 5.0.
858 19. Trivial typo in log message " temporarily refused connection" (the leading
861 20. If the log selector return_path_on_delivery was set and an address was
862 redirected to /dev/null, the delivery process crashed because it assumed
863 that a return path would always be set for a "successful" delivery. In this
864 case, the whole delivery is bypassed as an optimization, and therefore no
867 21. Internal re-arrangement: the function for sending a challenge and reading
868 a response while authentication was assuming a zero-terminated challenge
869 string. It's now changed to take a pointer and a length, to allow for
870 binary data in such strings.
872 22. Added the cyrus_sasl authenticator (code supplied by MBM).
874 23. Exim was not respecting finduser_retries when seeking the login of the
875 uid under which it was called; it was always trying 10 times. (The default
876 setting of finduser_retries is zero.) Also, it was sleeping after the final
877 failure, which is pointless.
879 24. Implemented tls_on_connect_ports.
881 25. Implemented acl_smtp_predata.
883 26. If the domain in control=submission is set empty, Exim assumes that the
884 authenticated id is a complete email address when it generates From: or
885 Sender: header lines.
887 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
888 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
889 chown and chgrp in /bin and hostname in /usr/bin.
891 28. Exim was keeping the "process log" file open after each use, just as it
892 does for the main log. This opens the possibility of it remaining open for
893 long periods when the USR1 signal hits a daemon. Occasional processlog
894 errors were reported, that could have been caused by this. Anyway, it seems
895 much more sensible not to leave this file open at all, so that is what now
898 29. The long-running daemon process does not normally write to the log once it
899 has entered its main loop, and it closes the log before doing so. This is
900 so that log files can straightforwardly be renamed and moved. However,
901 there are a couple of unusual error situations where the daemon does write
902 log entries, and I had neglected to close the log afterwards.
904 30. The text of an SMTP error response that was received during a remote
905 delivery was being truncated at 512 bytes. This is too short for some of
906 the long messages that one sometimes sees. I've increased the limit to
909 31. It is now possible to make retry rules that apply only when a message has a
910 specific sender, in particular, an empty sender.
912 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
913 it possible to be selective about when SMTP synchronization is enforced.
915 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
917 32. Implemented hosts_connection_nolog.
919 33. Added an ACL for QUIT.
921 34. Setting "delay_warning=" to disable warnings was not working; it gave a
924 35. Added mailbox_size and mailbox_filecount to appendfile.
926 36. Added control = no_multiline_responses to ACLs.
928 37. There was a bug in the logic of the code that waits for the clock to tick
929 in the case where the clock went backwards by a substantial amount such
930 that the microsecond fraction of "now" was more than the microsecond
931 fraction of "then" (but the whole seconds number was less).
933 38. Added support for the libradius Radius client library this is found on
934 FreeBSD (previously only the radiusclient library was supported).
940 1. When certain lookups returned multiple values in the form name=value, the
941 quoting of the values was not always being done properly. Specifically:
942 (a) If the value started with a double quote, but contained no whitespace,
944 (b) If the value contained whitespace other than a space character (i.e.
945 tabs or newlines or carriage returns) it was not quoted.
946 This fix has been applied to the mysql and pgsql lookups by writing a
947 separate quoting function and calling it from the lookup code. The fix
948 should probably also be applied to nisplus, ibase and oracle lookups, but
949 since I cannot test any of those, I have not disturbed their existing code.
951 2. A hit in the callout cache for a specific address caused a log line with no
952 reason for rejecting RCPT. Now it says "Previous (cached) callout
953 verification failure".
955 3. There was an off-by-one bug in the queryprogram router. An over-long
956 return line was truncated at 256 instead of 255 characters, thereby
957 overflowing its buffer with the terminating zero. As well as fixing this, I
958 have increased the buffer size to 1024 (and made a note to document this).
960 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
961 when Exim is waiting for an SMTP response from a remote server, Exim
962 restarts its select() call on the socket, thereby resetting its timeout.
963 This is not a problem when such interrupts are rare. Somebody set up a cron
964 job to run exiwhat every 2 minutes, which is less than the normal select()
965 timeout (5 or 10 minutes). This meant that the select() timeout never
966 kicked in because it was always reset. I have fixed this by comparing the
967 time when an interrupt arrives with the time at the start of the first call
968 to select(). If more time than the timeout has elapsed, the interrupt is
969 treated as a timeout.
971 5. Some internal re-factoring in preparation for the addition of Sieve
972 extensions (by MH). In particular, the "personal" test is moved to a
973 separate function, and given an option for scanning Cc: and Bcc: (which is
974 not set for Exim filters).
976 6. When Exim created an email address using the login of the caller as the
977 local part (e.g. when creating a From: or Sender: header line), it was not
978 quoting the local part when it contained special characters such as @.
980 7. Installed new OpenBSD configuration files.
982 8. Reworded some messages for syntax errors in "and" and "or" conditions to
983 try to make them clearer.
985 9. Callout options, other than the timeout value, were being ignored when
986 verifying sender addresses in header lines. For example, when using
988 verify = header_sender/callout=no_cache
990 the cache was (incorrectly) being used.
992 10. Added a missing instance of ${EXE} to the exim_install script; this affects
993 only the Cygwin environment.
995 11. When return_path_on_delivery was set as a log selector, if different remote
996 addresses in the same message used different return paths and parallel
997 remote delivery occurred, the wrong values would sometimes be logged.
998 (Whenever a remote delivery process finished, the return path value from
999 the most recently started remote delivery process was logged.)
1001 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1002 lines when AUTH and/or TLS are in use. This is the "received protocol"
1003 field. Exim used to use "asmtp" for authenticated SMTP, without any
1004 indication (in the protocol name) for TLS use. Now it follows the RFC and
1005 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1006 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1007 names appear in log lines as well as in Received: header lines.
1009 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1010 extensions, and comparison tests, and to fix some bugs.
1012 14. Changes to the "personal" filter test:
1014 (1) The test was buggy in that it was just doing the equivalent of
1015 "contains" tests on header lines. For example, if a user's address was
1016 anne@some.where, the "personal" test would incorrectly be true for
1018 To: susanne@some.where
1020 This test is now done by extracting each address from the header in turn,
1021 and checking the entire address. Other tests that are part of "personal"
1022 are now done using regular expressions (for example, to check local parts
1023 of addresses in From: header lines).
1025 (2) The list of non-personal local parts in From: addresses has been
1026 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1027 taken from the Sieve specification recommendations.
1029 (3) If the message contains any header line starting with "List-" it is
1030 treated as non-personal.
1032 (4) The test for "circular" in the Subject: header line has been removed
1033 because it now seems ill-conceived.
1035 15. Minor typos in src/EDITME comments corrected.
1037 16. Installed latest exipick from John Jetmore.
1039 17. If headers_add on a router specified a text string that was too long for
1040 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1041 of string_sprintf() is now avoided.
1043 18. $message_body_size was not set (it was always zero) when running the DATA
1044 ACL and the local_scan() function.
1046 19. For the "mail" command in an Exim filter, no default was being set for
1047 the once_repeat time, causing a random time value to be used if "once" was
1048 specified. (If the value happened to be <= 0, no repeat happened.) The
1049 default is now 0s, meaning "never repeat". The "vacation" command was OK
1050 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1051 (I found it when inspecting the code).
1053 20. There is now an overall timeout for performing a callout verification. It
1054 defaults to 4 times the callout timeout, which applies to individual SMTP
1055 commands during the callout. The overall timeout applies when there is more
1056 than one host that can be tried. The timeout is checked before trying the
1057 next host. This prevents very long delays if there are a large number of
1058 hosts and all are timing out (e.g. when the network connections are timing
1059 out). The value of the overall timeout can be changed by specifying an
1060 additional sub-option for "callout", called "maxwait". For example:
1062 verify = sender/callout=5s,maxwait=20s
1064 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1065 to the end before writing, but this should make it even safer).
1067 22. Exim was forgetting that it had advertised PIPELINING for the second and
1068 subsequent messages on an SMTP connection. It was also not resetting its
1069 memory on STARTTLS and an internal HELO.
1071 23. When Exim logs an SMTP synchronization error within a session, it now
1072 records whether PIPELINING has been advertised or not.
1074 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1075 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1076 rather than long int.
1078 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1080 26. Added the never_mail option to autoreply.
1086 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1087 crash if the getsockname() call failed; this can happen if a connection is
1088 closed very soon after it is established. The problem was simply in the
1089 order in which certain operations were done, causing Exim to try to write
1090 to the SMTP stream before it had set up the file descriptor. The bug has
1091 been fixed by making things happen in the correct order.
1097 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1098 before the connection was closed, thus losing the rejection response.
1100 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1101 some early Solaris releases, but causes trouble in current releases where
1102 socklen_t is defined.
1104 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1107 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1110 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1111 that releasing the top of it at the end releases what was used for sub-
1112 expansions (unless the block got too big). However, discard this block if
1113 the first thing is a variable or header, so that we can use its block when
1114 it is dynamic (useful for very large $message_headers, for example).
1116 6. Lookups now cache *every* query, not just the most recent. A new, separate
1117 store pool is used for this. It can be recovered when all lookup caches are
1118 flushed. Lookups now release memory at the end of their result strings.
1119 This has involved some general refactoring of the lookup sources.
1121 7. Some code has been added to the store_xxx() functions to reduce the amount
1122 of flapping under certain conditions.
1124 8. log_incoming_interface used to affect only the <= reception log lines. Now
1125 it causes the local interface and port to be added to several more SMTP log
1126 lines, for example "SMTP connection from", and rejection lines.
1128 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1130 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1132 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1133 could be overwritten at the end of the current message (or the start of a
1134 new message if it was set in a HELO ACL). The value is now preserved for
1135 the duration of the SMTP connection.
1137 12. If a transport had a headers_rewrite setting, and a matching header line
1138 contained an unqualified address, that address was qualified, even if it
1139 did not match any rewriting rules. The underlying bug was that the values
1140 of the flags that permit the existence of unqualified sender and recipient
1141 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1142 non-local messages, and by -bnq for local messages) were not being
1143 preserved with the message after it was received.
1145 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1146 "next input=" as part of the text comprising the host identity instead of
1147 the correct text. The code was using the same buffer for two different
1148 strings. However, depending on which order the printing function evaluated
1149 its arguments, the bug did not always show up. Under Linux, for example, my
1150 test suite worked just fine.
1152 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1153 doesn't work with some older versions of Perl. It has been changed to "my",
1154 which in any case is probably the better facility to use.
1156 15. A really picky compiler found some instances of statements for creating
1157 error messages that either had too many or two few arguments for the format
1160 16. The size of the buffer for calls to the DNS resolver has been increased
1161 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1162 for addresses that have a lot of PTR records. This alleviates a problem; it
1163 does not fully solve it.
1165 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1166 buffer now truncates the list and logs the incident, which is the same
1167 action as happens when Exim is looking up a host name and its aliases.
1168 Previously in this situation something unpredictable would happen;
1169 sometimes it was "internal error: store_reset failed".
1171 18. If a server dropped the connection unexpectedly when an Exim client was
1172 using GnuTLS and trying to read a response, the client delivery process
1173 crashed while trying to generate an error log message.
1175 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1176 string, for example:
1178 warn message = H1: something\nH2: something
1180 the text was added as a single header line from Exim's point of view
1181 though it ended up OK in the delivered message. However, searching for the
1182 second and subsequent header lines using $h_h2: did not work. This has been
1183 fixed. Similarly, if a system filter added multiple headers in this way,
1184 the routers could not see them.
1186 20. Expanded the error message when iplsearch is called with an invalid key to
1187 suggest using net-iplsearch in a host list.
1189 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1190 ACLs are no longer actually imposed (and a message to that effect is
1193 22. If a "gecos" field in a passwd entry contained escaped characters, in
1194 particular, if it contained a \" sequence, Exim got it wrong when building
1195 a From: or a Sender: header from that name. A second bug also caused
1196 incorrect handling when an unquoted " was present following a character
1197 that needed quoting.
1199 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1200 was not being matched caselessly.
1202 24. Arranged for all hyphens in the exim.8 source to be escaped with
1205 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1206 properly. Recipient callout cache records were still being keyed to include
1207 the sender, even when use_sender was set false. This led to far more
1208 callouts that were necessary. The sender is no longer included in the key
1209 when use_sender is false.
1211 26. Added "control = submission" modifier to ACLs.
1213 27. Added the ${base62d: operator to decode base 62 numbers.
1215 28. dnsdb lookups can now access SRV records.
1217 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1218 the configuration file.
1220 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1221 (-v) mode. This makes the output for a verbose queue run more intelligible.
1223 31. Added a use_postmaster feature to recipient callouts.
1225 32. Added the $body_zerocount variable, containing the number of binary zero
1226 bytes in the message body.
1228 33. The time of last modification of the "new" subdirectory is now used as the
1229 "mailbox time last read" when there is a quota error for a maildir
1232 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1234 35. Added +ignore_unknown as a special item in host lists.
1236 36. Code for decoding IPv6 addresses in host lists is now included, even if
1237 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1238 address was recognized as an IP address, but was then not correctly decoded
1239 into binary, causing unexpected and incorrect effects when compared with
1246 1. Very minor rewording of debugging text in manualroute to say "list of
1247 hosts" instead of "hostlist".
1249 2. If verify=header_syntax was set, and a header line with an unqualified
1250 address (no domain) and a large number of spaces between the end of the
1251 name and the colon was received, the reception process suffered a buffer
1252 overflow, and (when I tested it) crashed. This was caused by some obsolete
1253 code that should have been removed. The fix is to remove it!
1255 3. When running in the test harness, delay a bit after writing a bounce
1256 message to get a bit more predictability in the log output.
1258 4. Added a call to search_tidyup() just before forking a reception process. In
1259 theory, someone could use a lookup in the expansion of smtp_accept_max_
1260 per_host which, without the tidyup, could leave open a database connection.
1262 5. Added the variables $recipient_data and $sender_data which get set from a
1263 lookup success in an ACL "recipients" or "senders" condition, or a router
1264 "senders" option, similar to $domain_data and $local_part_data.
1266 6. Moved the writing of debug_print from before to after the "senders" test
1269 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1270 problems for message scanning, either using a data ACL, or using
1271 local_scan() because the Received: header was not generated till after they
1272 were called (in order to set the time as the time of reception completion).
1273 I have revised the way this works. The header is now generated after the
1274 body is received, but before the ACL or local_scan() are called. After they
1275 are run, the timestamp in the header is updated.
1281 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1282 before starting a queue runner without re-exec. This happened only when
1283 deliver_drop_privilege was set or when the Exim user was set to root. The
1284 effect of the bug was that timeouts during subsequent deliveries caused
1285 crashes instead of being properly handled. The handler is now left at its
1286 default (and expected) setting.
1288 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1289 message, again when deliver_drop_privilege is set or Exim is run as root.
1290 The bug described in (1) was not present in this case, but the tidying up
1291 of the other signals was missing. I have made the two cases consistent.
1293 3. The ignore_target_hosts setting on a manualroute router was being ignored
1294 for hosts that were looked up using the /MX notation.
1296 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1299 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1300 operated on the sender address. After changing the $sender_address to <>
1301 for the sender address verify, Exim was re-instated it as the original
1302 (before rewriting) address, but remembering that it had rewritten it, so it
1303 wasn't rewriting it again. This bug also had the effect of breaking the
1304 sender address verification caching when the sender address was rewritten.
1306 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1307 This has been changed so that if the ip literal address matches
1308 ignore_target_hosts, the router declines.
1310 7. Added expansion conditions match_domain, match_address, and match_local_
1311 part (NOT match_host).
1313 8. The placeholder for the Received: header didn't have a length field set.
1315 9. Added code to Exim itself and to exim_lock to test for a specific race
1316 condition that could lead to file corruption when using MBX delivery. The
1317 issue is with the lockfile that is created in /tmp. If this file is removed
1318 after a process has opened it but before that process has acquired a lock,
1319 there is the potential for a second process to recreate the file and also
1320 acquire a lock. This could lead to two Exim processes writing to the file
1321 at the same time. The added code performs the same test as UW imapd; it
1322 checks after acquiring the lock that its file descriptor still refers to
1323 the same named file.
1325 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1326 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1329 11. Added the smtp_active_hostname option. If used, this will typically be made
1330 to depend on the incoming interface address. Because $interface_address is
1331 not set up until the daemon has forked a reception process, error responses
1332 that can happen earlier (such as "too many connections") no longer contain
1335 12. If an expansion in a condition on a "warn" statement fails because a lookup
1336 defers, the "warn" statement is abandoned, and the next ACL statement is
1337 processed. Previously this caused the whole ACL to be aborted.
1339 13. Added the iplsearch lookup type.
1341 14. Added ident_timeout as a log selector.
1343 15. Added tls_certificate_verified as a log selector.
1345 16. Added a global option tls_require_ciphers (compare the smtp transport
1346 option of the same name). This controls incoming TLS connections.
1348 17. I finally figured out how to make tls_require_ciphers do a similar thing
1349 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1350 before starting the TLS session.
1352 18. Tabs are now shown as \t in -bP output.
1354 19. If the log selector return_path_on_delivery was set, Exim crashed when
1355 bouncing a message because it had too many Received: header lines.
1357 20. If two routers both had headers_remove settings, and the first one included
1358 a superfluous trailing colon, the final name in the first list and the
1359 first name in the second list were incorrectly joined into one item (with a
1360 colon in the middle).
1366 1. Added -C and -D options to the exinext utility, mainly to make it easier
1367 to include in the automated testing, but these could be helpful when
1368 multiple configurations are in use.
1370 2. The exinext utility was not formatting the output nicely when there was
1371 an alternate port involved in the retry record key, nor when there was a
1372 message id as well (for retries that were specific to a specific message
1373 and a specific host). It was also confused by IPv6 addresses, because of
1374 the additional colons they contain. I have fixed the IPv4 problem, and
1375 patched it up to do a reasonable job for IPv6.
1377 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1378 delivery, the log line now contains "pipelined" if PIPELINING was used.
1380 4. An SMTP transport process used to panic and die if the bind() call to set
1381 an explicit outgoing interface failed. This has been changed; it is now
1382 treated in the same way as a connect() failure.
1384 5. A reference to $sender_host_name in the part of a conditional expansion
1385 that was being skipped was still causing a DNS lookup. This no longer
1388 6. The def: expansion condition was not recognizing references to header lines
1389 that used bh_ and bheader_.
1391 7. Added the _cache feature to named lists.
1393 8. The code for checking quota_filecount in the appendfile transport was
1394 allowing one more file than it should have been.
1396 9. For compatibility with Sendmail, the command line option
1404 and sets the incoming protocol and host name (for trusted callers). The
1405 host name and its colon can be omitted when only the protocol is to be set.
1406 Note the Exim already has two private options, -pd and -ps, that refer to
1407 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1408 "s", but I don't think that's a major issue.
1410 10. A number of refactoring changes to the code, none of which should affect
1413 (a) The number of logging options was getting close to filling up the
1414 32-bit word that was used as a bit map. I have split them into two classes:
1415 those that are passed in the argument to log_write(), and those that are
1416 only ever tested independently outside of that function. These are now in
1417 separate 32-bit words, so there is plenty of room for expansion again.
1418 There is no change in the user interface or the logging behaviour.
1420 (b) When building, for example, log lines, the code previously used a
1421 macro that called string_cat() twice, in order to add two strings. This is
1422 not really sufficiently general. Furthermore, there was one instance where
1423 it was actually wrong because one of the argument was used twice, and in
1424 one call a function was used. (As it happened, calling the function twice
1425 did not affect the overall behaviour.) The macro has been replaced by a
1426 function that can join an arbitrary number of extra strings onto a growing
1429 (c) The code for expansion conditions now uses a table and a binary chop
1430 instead of a serial search (which was left over from when there were very
1431 few conditions). Also, it now recognizes conditions like "pam" even when
1432 the relevant support is not compiled in: a suitably worded error message is
1433 given if an attempt is made to use such a condition.
1435 11. Added ${time_interval:xxxxx}.
1437 12. A bug was causing one of the ddress fields not to be passed back correctly
1438 from remote delivery subprocesses. The field in question was not being
1439 subsequently used, so this caused to problems in practice.
1441 13. Added new log selectors queue_time and deliver_time.
1443 14. Might have fixed a bug in maildirsizefile handling that threw up
1444 "unexpected character" debug warnings, and recalculated the data
1445 unnecessarily. In any case, I expanded the warning message to give more
1448 15. Added the message "Restricted characters in address" to the statements in
1449 the default ACL that block characters like @ and % in local parts.
1451 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1452 Three changes have been made:
1454 (a) There was a serious bug; a negative response to MAIL caused the whole
1455 recipient domain to be cached as invalid, thereby blocking all messages
1456 to all local parts at the same domain, from all senders. This bug has
1457 been fixed. The domain is no longer cached after a negative response to
1458 MAIL if the sender used is not empty.
1460 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1463 (c) A new callout option, "use_sender" has been added for people who want
1464 the modified behaviour.
1470 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1473 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1474 indeed breaks things for older releases.
1476 3. Added additional logging to the case where there is a problem reading data
1477 from a filter that is running in a subprocess using a pipe, in order to
1478 try to track down a specific problem.
1480 4. Testing facility fudge: when running in the test harness and attempting
1481 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1482 getting "No route to host". Convert this to a timeout.
1484 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1487 6. Some OS don't have socklen_t but use size_t instead. This affects the
1488 fifth argument of getsockopt() amongst other things. This is now
1489 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1490 can be set for individual OS. I have set it for SunOS5, OSF1, and
1491 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1492 some earlier ones do not.
1494 7. Change 4.30/15 was not doing the test caselessly.
1496 8. The standard form for an IPv6 address literal was being rejected by address
1497 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1498 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1499 this, as well as the form without the "IPv6" on the front (but only when
1500 address literals are enabled, of course).
1502 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1504 10. Exim crashed if a message with an empty sender address specified by -f
1505 encountered a router with an errors_to setting. This could be provoked only
1506 by a command such as
1510 where an empty string was supplied; "<>" did not hit this bug.
1512 11. Installed PCRE release 4.5.
1514 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1515 remained set. It is now erased.
1517 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1518 times from message ids (which are base 36 rather than the normal 62).
1520 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1521 were being counted as actual protocol errors, and logged if the log
1522 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1523 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1524 and DATA following a set of rejected RCPTs do not count as protocol errors.
1525 In other words, Exim assumes they were pipelined, though this may not
1526 actually be the case. Of course, in all cases the client gets an
1527 appropriate error code.
1529 15. If a lookup fails in an ACL condition, a message about the failure may
1530 be available; it is used if testing the ACL cannot continue, because most
1531 such messages specify what the cause of the deferral is. However, some
1532 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1533 that caused an old message to be retained and used if a later statement
1534 caused a defer, replacing the real cause of the deferral.
1536 16. If an IP address had so many PTR records that the DNS lookup buffer
1537 was not large enough to hold them, Exim could crash while trying to process
1538 the truncated data. It now detects and logs this case.
1540 17. Further to 4.21/58, another change has been made: if (and only if) the
1541 first line of a message (the first header line) ends with CRLF, a bare LF
1542 in a subsequent header line has a space inserted after it, so as not to
1543 terminate the header.
1545 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1546 unnecessarily, used atoi() instead of strtol(), and didn't check the
1547 termination when getting file sizes from file names by regex.
1549 19. Completely re-implemented the support for maildirsize files, in the light
1550 of a number of problems with the previous contributed implementation
1551 (4.30/29). In particular:
1553 . If the quota is zero, the maildirsize file is maintained, but no quota is
1556 . If the maildir directory does not exist, it is created before any attempt
1557 to write a maildirsize file.
1559 . The quota value in the file is just a cache; if the quota is changed in
1560 the transport, the new value overrides.
1562 . A regular expression is available for excluding directories from the
1565 20. The autoreply transport checks the characters in options that define the
1566 message's headers; it allows continued headers, but it was checking with
1567 isspace() after an embedded newline instead of explicitly looking for a
1570 21. If all the "regular" hosts to which an address was routed had passed their
1571 expiry times, and had not reached their retry times, the address was
1572 bounced, even if fallback hosts were defined. Now Exim should go on to try
1575 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1576 equivalent code in the SMTP transport. Some hosts send humungous responses
1577 to HELO/EHLO, more than 1024 it seems.
1579 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1580 gives compiler warnings in some environments. I've now done it "properly",
1583 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1584 (because of problems with the built-in one) was declared to return uschar *
1585 instead of char *, causing compiler failure.
1587 25. Fixed a file descriptor leak when processing alias/forward files.
1589 26. Fixed a minor format string issue in dbfn.c.
1591 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1593 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1594 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1597 29. When a sender address is verified, it is cached, to save repeating the test
1598 when there is more than one recipient in a message. However, when the
1599 verification involves a callout, it is possible for different callout
1600 options to be set for different recipients. It is too complicated to keep
1601 track of this in the cache, so now Exim always runs a verification when a
1602 callout is required, relying on the callout cache for the optimization.
1603 The overhead is duplication of the address routing, but this should not be
1606 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1607 to be verified with callout=postmaster, and the main callout worked but the
1608 postmaster check failed, the verification correctly failed. However, if a
1609 subsequent RCPT command asked for sender verification *without* the
1610 postmaster check, incorrect caching caused this verification also to fail,
1613 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1614 it was not caching the DNS options (qualify_single, search_parents) that
1615 were used when the lookup failed. A subsequent lookup with different
1616 options therefore always gave the same answer, though there were cases
1617 where it should not have. (Example: a "domains = !$mx_any" option on a
1618 dnslookup router: the "domains" option is always processed without any
1619 widening, but the router might have qualify_single set.) Now Exim uses the
1620 cached value only when the same options are set.
1622 32. Added John Jetmore's "exipick" utility to the distribution.
1624 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1625 than a timeout (e.g. a certificate is required, and is not provided), an
1626 Exim server now closes the connection immediately. Previously it waited for
1627 the client to close - but if the client is SSL, it seems that they each
1628 wait for each other, leading to a delay before one of them times out.
1630 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1631 maintained 0.8.x compatibility because I don't think many are using it, and
1632 it is clearly obsolete.
1634 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1637 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1638 client certificate was expired. A simple patch fixes this, though I don't
1639 understand the full logic of why the verify callback is called multiple
1642 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1643 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1644 which causes problems with some clients (such as the Certicom SSL Plus
1645 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1646 disables the coutermeasure allowing Eudora to connect."
1648 38. Exim was not checking that a write() to a log file succeeded. This could
1649 lead to Bad Things if a log got too big, in particular if it hit a file
1650 size limit. Exim now panics and dies if it cannot write to a log file, just
1651 as it does if it cannot open a log file.
1653 39. Modified OS/Makefile-Linux so that it now contains
1655 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1657 The two -D definitions ensure that Exim is compiled with large file
1658 support, which makes it possible to handle log files that are bigger than
1661 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1662 instance) a domain was checked against a named list that involved a lookup,
1663 causing $domain_data to be set, then another domain was checked against the
1664 same list, then the first domain was re-checked, the value of $domain_data
1665 after the final check could be wrong. In particular, if the second check
1666 failed, it could be set empty. This bug probably also applied to
1669 41. The strip_trailing_dot option was not being applied to the address given
1670 with the -f command-line option.
1672 42. The code for reading a message's header from the spool was incrementing
1673 $received_count, but never initializing it. This meant that the value was
1674 incorrect (doubled) while delivering a message in the same process in which
1675 it was received. In the most common configuration of Exim, this never
1676 happens - a fresh exec is done - but it can happen when
1677 deliver_drop_privilege is set.
1679 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1680 it now includes up to 150 characters of the unexpected data in the log
1683 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1684 and building data strings. The size of both of these buffers was 10 000
1685 bytes - far larger than anybody would *ever* want, thought I. Needless to
1686 say, somebody hit the limit. I have increased the maximum line length to
1687 20 000 and the maximum data length of concatenated lines to 100 000. I have
1688 also fixed two bugs, because there was no checking on these buffers. Tsk,
1689 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1690 buffer is too small.
1692 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1693 lsearch lookups. Now it does.
1695 46. When parsing a route_list item in a manualroute router, a fixed-length
1696 buffer was used for the list of hosts. I made this 1024 bytes long,
1697 thinking that nobody would ever have a list of hosts that long. Wrong.
1698 Somebody had a whole pile of complicated expansion conditions, and the
1699 string was silently truncated, leading to an expansion error. It turns out
1700 that it is easier to change to an unlimited length (owing to other changes
1701 that have happened since this code was originally written) than to build
1702 structure for giving a limitation error. The length of the item that
1703 expands into the list of hosts is now unlimited.
1705 47. The lsearch lookup could not handle data where the length of text line was
1706 more than 4095 characters. Such lines were truncated, leading to shortened
1707 data being returned. It should now handle lines of any length.
1709 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1710 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1713 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1714 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1715 no longer happens in comment lines. A list of replacements is now placed
1716 at the head of all of the source files, except those whose only change is
1717 to replace PERL_COMMAND in the very first #! line.
1719 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1720 messages on the queue, with a bottom-up merge sort, using code contributed
1721 by Michael Haardt. This should make operations like -bp somewhat faster on
1722 large queues. It won't affect queue runners, except when queue_run_in_order
1725 51. Installed eximstats 1.31 in the distribution.
1727 52. Added support for SRV lookups to the dnslookup router.
1729 53. If an ACL referred to $message_body or $message_body_end, the value was not
1730 reset for any messages that followed in the same SMTP session.
1732 54. The store-handling optimization for building very long strings was not
1733 differentiating between the different store pools. I don't think this
1734 actually made any difference in practice, but I've tidied it.
1736 55. While running the routers to verify a sender address, $sender_address
1737 was still set to the sender address. This is wrong, because when routing to
1738 send a bounce to the sender, it would be empty. Therefore, I have changed
1739 it so that, while verifying a sender address, $sender_address is set to <>.
1740 (There is no change to what happens when verifying a recipient address.)
1742 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1743 target A or AAAA records (if not already returned) without resetting the
1744 qualify_single or search_parents options of the DNS resolver. These are
1745 inappropriate in this case because the targets of MX and SRV records must
1746 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1747 target that, when qualified, matched something in the local domain. These
1748 two options are now turned off when doing these lookups.
1750 57. It seems that at least some releases of Reiserfs (which does not have the
1751 concept of a fixed number of inodes) returns zero and not -1 for the
1752 number of available inodes. This interacted badly with check_spool_inodes,
1753 which assumed that -1 was the "no such thing" setting. What I have done is
1754 to check that the total number of inodes is greater than zero before doing
1755 the test of how many are available.
1757 58. When a "warn" ACL statement has a log_message modifier, the message is
1758 remembered, and not repeated. This is to avoid a lot of repetition when a
1759 message has many recipients that cause the same warning to be written.
1760 Howewer, Exim was preserving the list of already written lines for an
1761 entire SMTP session, which doesn't seem right. The memory is now reset if a
1762 new message is started.
1764 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1765 debugging output unless log_rewrite was also set.
1767 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1768 of (int)(handle) when we know that handle contains (void *)(-1).
1770 61. The Exim daemon panic-logs an error return when it closes the incoming
1771 connection. However "connection reset by peer" seems to be common, and
1772 isn't really an error worthy of noting specially, so that particular error
1775 62. When Exim is trying to find all the local interfaces, it used to panic and
1776 die if the ioctl to get the interface flags failed. However, it seems that
1777 on at least one OS (Solaris 9) it is possible to have an interface that is
1778 included in the list of interfaces, but for which you get a failure error
1779 for this call. This happens when the interface is not "plumbed" into a
1780 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1781 failure of the "get flags" call assumes that the interface is down.
1783 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1784 makes life easier for people who are doing arithmetic on fields extracted
1785 from dates, where you often get leading zeros that should not be
1786 interpreted as octal.
1788 64. Added qualify_domain to the redirect router, to override the global
1791 65. If a pathologically long header line contained very many addresses (the
1792 report of this problem mentioned 10 000) and each of them was rewritten,
1793 Exim could use up a very large amount of memory. (It kept on making new
1794 copies of the header line as it rewrote, and never released the old ones.)
1795 At the expense of a bit more processing, the header rewriting function has
1796 been changed so that it no longer eats memory in this way.
1798 66. The generation of the Received: header has been moved from the time that a
1799 message starts to be received, to the time that it finishes. The timestamp
1800 in the Received: header should now be very close to that of the <= log
1801 line. There are two side-effects of this change:
1803 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1804 logged header lines no longer include the local Received: line, because
1805 it has not yet been created. The same applies to a copy of the message
1806 that is returned to a non-SMTP sender when a message is rejected.
1808 (b) When a filter file is tested using -bf, no additional Received: header
1809 is added to the test message. After some thought, I decided that this
1812 This change does not affect the value of $received_for. It is still set
1813 after address rewriting, but before local_scan() is called.
1815 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1817 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1818 gave an unhelpful panic error message, and a defer error. I have managed to
1819 change this behaviour so that it now rejects any supplied certificate,
1820 which seems right, as the list of acceptable certificates is empty.
1822 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1823 gave an unhelpful defer error. I have not managed to make this reject any
1824 supplied certificates, but the error message it gives is "no certificate
1825 supplied", which is not helpful.
1827 70. exigrep's output now also includes lines that are not associated with any
1828 message, but which match the given pattern. Implemented by a patch from
1829 Martin Sluka, which also tidied up the Perl a bit.
1831 71. Recipient callout verification, like sender verification, was using <> in
1832 the MAIL FROM command. This isn't really the right thing, since the actual
1833 sender may affect whether the remote host accepts the recipient or not. I
1834 have changed it to use the actual sender in the callout; this means that
1835 the cache record is now keyed on a recipient/sender pair, not just the
1836 recipient address. There doesn't seem to be a real danger of callout loops,
1837 since a callout by the remote host to check the sender would use <>.
1838 [SEE ABOVE: changed after hitting problems.]
1840 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1841 temporary errors. However, in the case of such a code being given after
1842 the end of a data transmission (i.e. after ".") Exim was failing to write
1843 a retry record for the message. (Yes, there was some broken host that was
1844 actually sending 8xx at this point.)
1846 73. An unknown lookup type in a host list could cause Exim to panic-die when
1847 the list was checked. (An example that provoked this was putting <; in the
1848 middle of a list instead of at the start.) If this happened during a DATA
1849 ACL check, a -D file could be left lying around. This kind of configuration
1850 error no longer causes Exim to die; instead it causes a defer errror. The
1851 incident is still logged to the main and panic logs.
1853 74. Buglet left over from Exim 3 conversion. The message "too many messages
1854 in one connection" was written to the rejectlog but not the mainlog, except
1855 when address rewriting (yes!) was being logged.
1857 75. Added write_rejectlog option.
1859 76. When a system filter was run not as root (that is, when system_filter_user
1860 was set), the values of the $n variables were not being returned to the
1861 main process; thus, they were not subsequently available in the $sn
1864 77. Added +return_path_on_delivery log selector.
1866 78. A connection timeout was being treated differently from recipients deferred
1867 when testing hosts_max_try with a message that was older than the host's
1868 retry timeout. (The host should not be counted, thus allowing all hosts to
1869 be tried at least once before bouncing.) This may have been the cause of an
1870 occasionally reported bug whereby a message would remain on the queue
1871 longer than the retry timeout, but would be bounced if a delivery was
1872 forced. I say "may" because I never totally pinned down the problem;
1873 setting up timeout/retry tests is difficult. See also the next item.
1875 79. The ultimate address timeout was not being applied to errors that involved
1876 a combination of host plus message (for example, a timeout on a MAIL
1877 command). When an address resolved to a number of possible hosts, and they
1878 were not all tried for each delivery (e.g. because of hosts_max_try), a
1879 message could remain on the queue longer than the retry timeout.
1881 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1884 81. Fixed an obscure SMTP outgoing bug which required at least the following
1885 conditions: (a) there was another message waiting for the same server;
1886 (b) the server returned 5xx to all RCPT commands in the first message so
1887 that the message was not completed; (c) the server dropped the connection
1888 or gave a negative response to the RSET that Exim sends to abort the
1889 transaction. The observed case was a dropped connection after DATA that had
1890 been sent in pipelining mode. That is, the server had advertised PIPELINING
1891 but was not implementing it correctly. The effect of the bug was incorrect
1892 behaviour, such as trying another host, and this could lead to a crash.
1898 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1899 and daemon.c were passed as pointers to ints; they should have been
1900 pointers to socklen_t variables (which are typically unsigned ints).
1902 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1905 3. Fixed a really odd bug that affected only the testing scheme; patching a
1906 certain fixed string in the binary changed the value of another string that
1907 happened to be identical to the end of the original first string.
1909 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1910 name", it returns that address as the IP address. On some operating
1911 systems (e.g. Solaris), it also passes back the IP address string as the
1912 "host name". However, on others (e.g. Linux), it passes back an empty
1913 string. Exim wasn't checking for this, and was changing the host name to an
1914 empty string, assuming it had been canonicized.
1916 5. Although rare, it is permitted to have more than one PTR record for a given
1917 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1918 all the names associated with an address, because they do in Solaris.
1919 However, it seems that they do not in Linux for data that comes from the
1920 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1921 I found this out when I moved to a new Linux workstation and tried to run
1922 the Exim test suite.
1924 To get round this problem I have changed the code so that it now does its
1925 own call to the DNS to look up PTR records when searching for a host name.
1926 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1927 addresses that are only in /etc/hosts are still found.
1929 This behaviour is, however, controlled by an option called host_lookup_
1930 order, which defaults to "bydns:byaddr". If people want to use the other
1931 order, or indeed, just use one or the other means of lookup, they can
1932 specify it in this variable.
1934 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1935 some operating systems, this comes back from gethostbyaddr() as an empty
1936 string, and this is what Exim used to test for. However, it seems that in
1937 other systems, "." is yielded. Exim now tests for this case too.
1939 7. The values of check_spool_space and check_log_space are now held internally
1940 as a number of kilobytes instead of an absolute number of bytes. If a
1941 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1942 kilobyte. This means that much larger values can be stored.
1944 8. Exim monitor: an attempt to get the action menu when not actually pointing
1945 at a message produces an empty menu entitled "No message selected". This
1946 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1947 no entries in it ("Shell widget menu has zero width and/or height"). So I
1948 have added a single, blank menu entry in this case.
1950 9. Added ${quote_local_part.
1952 10. MIME decoding is now applied to the contents of Subject: header lines when
1955 11. Now that a reference to $sender_host_address automatically causes a reverse
1956 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1957 host lookup before query-style lookups in lists that might use this
1958 variable. This has therefore been abolished, and the "net-" prefix is no
1959 longer necessary for query-style lookups.
1961 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1962 have been a typo for LFLAGS, so it has been changed.
1964 13. The install script calls Exim with "-C /dev/null" in order to find the
1965 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1966 to be output. Howeve, since Exim outputs its version number before the
1967 error, it didn't break the script. It just looked ugly. I fixed this by
1968 always allowing "-C /dev/null" if the caller is root.
1970 14. Ignore overlarge ACL variable number when reading spool file - insurance
1971 against a later release with more variables having written the file.
1973 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1974 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1975 this, as well as the form without the "IPv6" on the front.
1977 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1978 OS/Makefile-Darwin file.
1980 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1981 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1983 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1984 variable. However, some OS use ints for inodes. I've added cast to long int
1985 to get rid of the compiler warning.
1987 19. I had forgotten to lock out "/../" in configuration file names when
1988 ALT_CONFIG_PREFIX was set.
1990 20. Routers used for verification do not need to specify transports. However,
1991 if such a router generated a host list, and callout was configured, Exim
1992 crashed, because it could not find a port number from the (non-existent)
1993 transport. It now assumes port 25 in this circumstance.
1995 21. Added the -t option to exigrep.
1997 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1998 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1999 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2000 them is set, LOOKUP_LSEARCH is forced.
2002 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2004 24. Added sender and host information to the "rejected by local_scan()" log
2005 line; previously there was no indication of these.
2007 25. Added .include_if_exists.
2009 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2010 Linux. It turns out that not all file systems support this. Apparently some
2011 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2012 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2013 not supported on the file descriptor, is now ignored when Exim is trying to
2014 sync a directory. This applies only to Linux.
2016 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2018 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2019 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2022 29. Added support for maildirsize files from supplied patch (modified a bit).
2024 30. The use of :fail: followed by an empty string could lead Exim to respond to
2025 sender verification failures with (e.g.):
2027 550 Verification failed for <xxx>
2028 550 Sender verify failed
2030 where the first response line was missing the '-' that indicates it is not
2031 the final line of the response.
2033 31. The loop for finding the name of the user that called Exim had a hardwired
2034 limit of 10; it now uses the value of finduser_retries, which is used for
2035 all other user lookups.
2037 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2040 33. Exim was neglecting to zero errno before one call of strtol() when
2041 expanding a string and expecting an integer value. On some systems this
2042 resulted in spurious "integer overflow" errors. Also, it was casting the
2043 result into an int without checking.
2045 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2046 did not work. The code looks as if it has *never* worked, though it appears
2047 to have been documented since at least releast 1.62. I have made it work.
2049 35. The "timeout_DNS" error in retry rules, also documented since at least
2050 1.62, also never worked. As it isn't clear exactly what this means, and
2051 clearly it isn't a major issue, I have abolished the feature by treating it
2052 as "timeout", and writing a warning to the main and panic logs.
2054 36. The display of retry rules for -brt wasn't always showing the error code
2057 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2058 timeout_connect_A, timeout_connect_MX.
2060 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2061 to the empty sender.
2063 39. The daemon was not analysing the content of -oX till after it had closed
2064 stderr and disconnected from the controlling terminal. This meant that any
2065 syntax errors were only noted on the panic log, and the return code from
2066 the command was 0. By re-arranging the code a little, I've made the
2067 decoding happen first, so such errors now appear on stderr, and the return
2068 code is 1. However, the actual setting up of the sockets still happens in
2069 the disconnected process, so errors there are still only recorded on the
2072 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2073 connections (as happens on some IP stacks) was logged at start up time as
2074 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2075 it from "IPv6 and IPv4", which means that two separate sockets are being
2078 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2079 says whether AF_INET or AF_INET6 was passed as an argument.
2081 42. Exiwhat output was messed up when time zones were included in log
2084 43. Exiwhat now gives more information about the daemon's listening ports,
2085 and whether -tls-on-connect was used.
2087 44. The "port" option of the smtp transport is now expanded.
2089 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2090 silently ignored. Now an error message is written to the main and panic
2093 46. There's a new ACL modifier called "logwrite" which writes to a log file
2094 as soon as it is encountered.
2096 47. Added $local_user_uid and $local_user_gid at routing time.
2098 48. Exim crashed when trying to verify a sender address that was being
2101 49. Exim was recognizing only a space character after ".include". It now also
2102 recognizes a tab character.
2104 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2105 extracting the relevant information from the specification. The man page no
2106 longer contains scrambled data for the -d option, and I've added a section
2107 at the front about calling Exim under different names.
2109 51. Added "extra_headers" argument to the "mail" command in filter files.
2111 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2114 53. Installed eximstats 1.29.
2116 54. Added transport_filter_timeout as a generic transport option.
2118 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2119 Cc: header lines. This was required by RFC 822, but it not required by RFC
2122 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2123 incoming messages that did not have them. Now it does so only if the
2124 message originates locally, that is, if there is no associated remote host
2125 address. When Resent- header lines are present, this applies to the Resent-
2126 lines rather than the non-Resent- lines.
2128 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2129 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2131 58. Messages for configuration errors now include the name of the main
2132 configuration file - useful now that there may be more than one file in a
2133 list (.included file names were always shown).
2135 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2136 for those rare installations that do not start the daemon as root or run it
2137 setuid root. I've cut out the call to initgroups() if the daemon is not
2140 60. The Exim user and group can now be bound into the binary as text strings
2141 that are looked up at the start of Exim's processing.
2143 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2145 62. Added $mailstore_basename variable.
2147 63. Installed patch to sieve.c from Michael Haardt.
2149 64. When Exim failed to open the panic log after failing to open the main log,
2150 the original message it was trying to log was written to stderr and debug
2151 output, but if they were not available (the usual case in production), it
2152 was lost. Now it is written to syslog before the two lines that record the
2153 failures to open the logs.
2155 65. Users' Exim filters run in subprocesses under the user's uid. It is
2156 possible for a "deliver" command or an alias in a "personal" command to
2157 provoke an address rewrite. If logging of address rewriting is configured,
2158 this fails because the process is not running as root or exim. There may be
2159 a better way of dealing with this, but for the moment (because 4.30 needs
2160 to be released), I have disabled address rewrite logging when running a
2161 filter in a non-root, non-exim process.
2167 1. The buildconfig auxiliary program wasn't quoting the value set for
2168 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2169 not defined. This bug was masked in 4.22 by the effect that was fixed in
2172 2. Some messages that were rejected after a message id was allocated were
2173 shown as "incomplete" by exigrep. It no longer does this for messages that
2174 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2176 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2177 have allow_domain_literals set, the ID did not get logged in the <= line.
2178 Domain literals are now always recognized in Message-ID: header lines.
2180 4. The first argument for a ${extract expansion item is the key name or field
2181 number. Leading and trailing spaces in this item were not being ignored,
2182 causing some misleading effects.
2184 5. When deliver_drop_privilege was set, single queue runner processes started
2185 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2186 same command in the process it spins off) were not dropping privilege.
2188 6. When the daemon running as "exim" started a queue runner, it always
2189 re-executed Exim in the spun-off process. This is a waste of effort when
2190 deliver_drop_privilege is set. The new process now just calls the
2191 queue-runner function directly.
2197 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2200 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2201 ignored. Though the use of -oP was forcing the writing of a pid file, it
2202 was always written to the default place.
2204 3. If the message "no IP address found for host xxxx" is generated during
2205 incoming verification, it is now followed by identification of the incoming
2206 connection (so you can more easily find what provoked it).
2208 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2210 5. Added some features to "harden" Exim a bit more against certain attacks:
2212 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2213 be put in Local/Makefile. This is like the never_users runtime option,
2214 but it cannot be overridden. The default setting is "root".
2216 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2217 prefix string with which any file named in a -C command line option
2220 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2221 is retained for -C and -D only if the caller of Exim is root. Without
2222 it, the exim user may also use -C and -D and retain privilege.
2224 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2225 command line option is disabled.
2227 6. Macro names set by the -D option must start with an upper case letter, just
2228 like macro names defined in the configuration file.
2230 7. Added "dereference=" facility to LDAP.
2232 8. Two instances of the typo "uknown" in the source files are fixed.
2234 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2235 the Configure-Makefile script screwed up while processing it.
2237 10. Incorporated PCRE 4.4.
2239 11. The SMTP synchronization check was not operating right at the start of an
2240 SMTP session. For example, it could not catch a HELO sent before the client
2241 waited for the greeting. There is now a check for outstanding input at the
2242 point when the greeting is written. Because of the duplex, asynchronous
2243 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2244 way, but not yet received, when the check is performed.
2246 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2247 on TCP/IP sockets, because this apparently causes some broken clients to
2250 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2251 unchanged) from the Cygwin maintainer.
2253 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2254 was supported instead of testing for mbx. Effectively a typo.
2256 15. The spa authenticator server code was not checking that the input it
2257 received was valid base64.
2259 16. The debug output line for the "set" modifier in ACLs was not showing the
2260 name of the variable that was being set.
2262 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2264 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2265 lookup on its own. Something else was needed to trigger the lookup. For
2266 example, a match in host_lookup or the need for a host name in a host list.
2267 Now, if $sender_host_name is referenced and the host name has not yet been
2268 looked up, a lookup is performed. If the lookup fails, the variable remains
2269 empty, and $host_lookup_failed is set to "1".
2271 19. Added "eqi" as a case-independent comparison operator.
2273 20. The saslauthd authentication condition could segfault if neither service
2274 nor realm was specified.
2276 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2277 error message that was logged was misleading, and incoming SMTP
2278 connections were dropped. The message is now more accurate, and temporary
2279 errors are given to SMTP connections.
2281 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2282 (except RSET) with a 421 error, until QUIT is received. However, it was
2283 failing to send a response to QUIT.
2285 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2286 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2287 ACL is now run after the helo_try_verify_hosts code.
2289 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2290 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2291 case-independent, so other case variants are also recognized). Apparently
2292 some systems use these upper case variants.
2294 25. If more than two messages were waiting for the same host, and a transport
2295 filter was specified for the transport, Exim sent two messages over the
2296 same TCP/IP connection, and then failed with "socket operation on non-
2297 socket" when it tried to send the third.
2299 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2301 27. The extern definition of crypt16() in expand.c was not being excluded when
2302 the OS had its own crypt16() function.
2304 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2305 as a preferred synonym for return_size_limit, both as an option and as an
2308 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2310 30. Changed the default configuration ACL to relax the local part checking rule
2311 for addresses that are not in any local domains. For these addresses,
2312 slashes and pipe symbols are allowed within local parts, but the sequence
2313 /../ is explicitly forbidden.
2315 31. SPA server authentication was not clearing the challenge buffer before
2318 32. log_message in a "warn" ACL statement was writing to the reject log as
2319 well as to the main log, which contradicts the documentation and doesn't
2320 seem right (because no rejection is happening). So I have stopped it.
2322 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2323 However, I am unable to do any testing of this.
2325 34. Fixed an infelicity in the appendfile transport. When checking directories
2326 for a mailbox, to see if any needed to be created, it was accidentally
2327 using path names with one or more superfluous leading slashes; tracing
2328 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2330 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2331 contents are added to the log line that is written for every discarded
2332 recipient. (Previously a log_message setting was ignored.)
2334 36. The ${quote: operator now quotes the string if it is empty.
2336 37. The install script runs exim in order to find its version number. If for
2337 some reason other than non-existence or emptiness, which it checks, it
2338 could not run './exim', it was installing it with an empty version number,
2339 i.e. as "exim-". This error state is now caught, and the installation is
2342 38. An argument was missing from the function that creates an error message
2343 when Exim fails to connect to the socket for saslauthd authentication.
2344 This could cause Exim to crash, or give a corrupted message.
2346 39. Added isip, isip4, and isip6 to ${if conditions.
2348 40. The ACL variables $acl_xx are now saved with the message, and can be
2349 accessed later in routers, transports, and filters.
2351 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2352 strings in the file are not string-expanded.
2354 42. If a MAIL command specified a SIZE value that was too large to fit into an
2355 int variable, the check against message_size_limit failed. Such values are
2356 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2357 day this will have to be increased, but I don't think I want to be around
2358 when emails are that large.
2365 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2366 iconv() is not standard in FreeBSD.
2368 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2369 IPv6 enabled. The observed symptom was a segmentation fault on return from
2370 the function os_common_find_running_interfaces() in src/os.c.
2372 3. In the check_special_case() function in daemon.c I had used "errno" as an
2373 argument name, which causes warnings on some systems. This was basically a
2374 typo, since it was named "eno" in the comments!
2376 4. The code that waits for the clock to tick (at a resolution of some fraction
2377 of a second) so as to ensure message-id uniqueness was always waiting for
2378 at least one whole tick, when it could have waited for less. [This is
2379 almost certainly not relevant at current processor speeds, where it is
2380 unlikely to ever wait at all. But we try to future-proof.]
2382 5. The function that sleeps for a time interval that includes fractions of a
2383 second contained a race. It did not block SIGALRM between setting the
2384 timer, and suspending (a couple of lines later). If the interval was short
2385 and the sigsuspend() was delayed until after it had expired, the suspension
2386 never ended. On busy systems this could lead to processes getting stuck for
2389 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2390 process, before it forks any delivery processes. The open lookup caching
2391 mechanism meant that the open file or database connection was passed into
2392 the delivery process. The problem was that delivery processes always tidy
2393 up cached lookup data. This could cause a problem for the next delivery
2394 process started by the queue runner, because the external queue runner
2395 process does not know about the closure. So the next delivery process
2396 still has data in the lookup cache. In the case of a file lookup, there was
2397 no problem because closing a file descriptor in a subprocess doesn't affect
2398 the parent. However, if the lookup was caching a connection to a database,
2399 the connection was closed, and the second delivery process was likely to
2400 see errors such as "PGSQL: query failed: server closed the connection
2401 unexpectedly". The problem has been fixed by closing all cached lookups
2402 in a queue runner before running a delivery process.
2404 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2405 seem to have the "const" qualifier which it has on other OS. I've
2408 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2409 *of the same type* (client or server) with the same public name that an
2410 error should be diagnosed.
2412 9. When Exim looked up a host name for an IP address, but failed to find the
2413 original IP address when looking up the host name (a safety check), it
2414 output the message "<ip address> does not match any IP for NULL", which was
2415 confusing, to say the least. The bug was that the host name should have
2416 appeared instead of "NULL".
2418 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2419 user that is built into the binary, and the -C or -D options is used, root
2420 privilege is dropped before the configuration file is read. In addition,
2421 logging is switched to stderr instead of the normal log files. If the
2422 configuration then re-defines the Exim user, the unprivileged environment
2423 is probably not what is expected, so Exim logs a panic warning message (but
2426 However, if deliver_drop_privilege is set, the unprivileged state may well
2427 be exactly what is intended, so the warning has been cut out in that case,
2428 and Exim is allowed to try to write to its normal log files.
2434 1. smtp_return_error_details was not giving details for temporary sender
2435 or receiver verification errors.
2437 2. Diagnose a configuration error if two authenticators have the same public
2440 3. Exim used not to create the message log file for a message until the first
2441 delivery attempt. This could be confusing when incoming messages were held
2442 for policy or load reasons. The message log file is now created at the time
2443 the message is received, and an initial "Received" line is written to it.
2445 4. The automatically generated man page for command line options had a minor
2446 bug that caused no ill effects; however, a more serious problem was that
2447 the procedure for building the man page automatically didn't always
2448 operate. Consequently, release 4.20 contains an out-of-date version. This
2449 shouldn't happen again.
2451 5. When building Exim with embedded Perl support, the script that builds the
2452 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2453 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2455 6. The freeze_tell option was not being used for messages that were frozen on
2456 arrival, either by an ACL or by local_scan().
2458 7. Added the smtp_incomplete_transaction log selector.
2460 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2461 was accepting AUTH without a new EHLO.
2463 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2464 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2467 10. Logging of TCP/IP connections (when configured) now happens in the main
2468 daemon process instead of the child process, so that the TCP/IP connection
2469 count is more accurate (but it can never be perfect).
2471 11. The use of "drop" in a nested ACL was not being handled correctly in the
2472 outer ACL. Now, if condition failure induced by the nested "drop" causes
2473 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2474 or "require"), the connection is dropped.
2476 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2477 that yield "discard" can now be used with an "accept" or a "discard" verb,
2478 but an error is generated for any others (because I can't see a useful way
2479 to define what should happen).
2481 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2482 are now processed in the same way as lines in the Exim configuration file.
2483 In particular, continuation lines are supported.
2485 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2487 15. Added -ti meaning -t -i.
2489 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2490 domains, and warn by logging if others are found.
2492 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2493 returned from GIFCONF when Exim is trying to find the list of interfaces on
2494 a host. The code in os.c has been modified to copy each ifreq to an aligned
2495 structure in all cases.
2497 Also, in some cases, the returned ifreq's were being copied to a 'struct
2498 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2499 means the last couple of bytes of an IPv6 address could be chopped if the
2500 ifreq contained only a normal sockaddr (14 bytes storage).
2502 18. Named domain lists were not supported in the hosts_treat_as_local option.
2503 An entry such as +xxxx was not recognized, and was treated as a literal
2506 19. Ensure that header lines added by a DATA ACL are included in the reject log
2507 if the ACL subsequently rejects the message.
2509 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2510 MD5 (which is deprecated).
2512 21. When testing a filter file using -bf, Exim was writing a message when it
2513 took the sender from a "From " line in the message, but it was not doing so
2514 when it took $return_path from a Return-Path: header line. It now does.
2516 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2517 with a valid header line field name (a series of printing characters
2518 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2520 23. Changed "disc" in the source to "disk" to conform to the documentation and
2521 the book and for uniformity.
2523 24. Ignore Sendmail's -Ooption=value command line item.
2525 25. When execve() failed while trying to run a command in a pipe transport,
2526 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2527 could be confused with a return value of 69 from the command itself. This
2528 has been changed to 127, the value the shell returns if it is asked to run
2529 a non-existent command. The wording for the related log line suggests a
2530 non-existent command as the problem.
2532 26. If received_header_text expands to an empty string, do not add a Received:
2533 header line to the message. (Well, it adds a token one on the spool, but
2534 marks it "old" so that it doesn't get used or transmitted.)
2536 27. Installed eximstats 1.28 (addition of -nt option).
2538 28. There was no check for failure on the call to getsockname() in the daemon
2539 code. This can fail if there is a shortage of resources on the system, with
2540 ENOMEM, for example. A temporary error is now given on failure.
2542 29. Contrary to the C standard, it seems that in some environments, the
2543 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2544 program. Exim now does this explicitly; it affects the formatting of
2545 timestamps using strftime().
2547 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2548 complaints. I've now initialized all the variables, to avoid this.
2550 32. Header lines added by a system filter were not being "seen" during
2551 transport-time rewrites.
2553 33. The info_callback() function passed to OpenSSL is set up with type void
2554 (*)(SSL *, int, int), as described somewhere. However, when calling the
2555 function (actually a macro) that sets it up, the type void(*)() is
2556 expected. I've put in a cast to prevent warnings from picky compilers.
2558 34. If a DNS black list lookup found a CNAME record, but there were no A
2559 records associated with the domain it pointed at, Exim crashed.
2561 35. If a DNS black list lookup returned more than one A record, Exim ignored
2562 all but the first. It now scans all returned addresses if a particular IP
2563 value is being sought. In this situation, the contents of the
2564 $dnslist_value variable are a list of all the addresses, separated by a
2567 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2568 to accept a host name and all its aliases if the forward lookup for any of
2569 them yielded the IP address of the incoming connection. Now it accepts only
2570 those names whose forward lookup yields the correct IP address. Any other
2571 names are discarded. This closes a loophole whereby a rogue DNS
2572 administrator could create reverse DNS records to break through a
2573 wildcarded host restriction in an ACL.
2575 37. If a user filter or a system filter that ran in a subprocess used any of
2576 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2577 the wrong values were passed to the pipe command ($thisaddress had the
2578 value of $0, $0 had the value of $1, etc). This bug was introduced by
2579 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2581 38. Improved the line breaking for long SMTP error messages from ACLs.
2582 Previously, if there was no break point between 40 and 75 characters, Exim
2583 left the rest of the message alone. Two changes have been made: (a) I've
2584 reduced the minimum length to 35 characters; (b) if it can't find a break
2585 point between 35 and 75 characters, it looks ahead and uses the first one
2586 that it finds. This may give the occasional overlong line, but at least the
2587 remaining text gets split now.
2589 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2590 file descriptors might be low, and that setting 1000 would always raise it.
2591 It turns out that in some environments, the limit is already over 1000 and
2592 that lowering it causes trouble. So now Exim takes care not to decrease it.
2594 40. When delivering a message, the value of $return_path is set to $sender_
2595 address at the start of routing (routers may change the value). By an
2596 oversight, this default was not being set up when an address was tested by
2597 -bt or -bv, which affected the outcome if any router or filter referred to
2600 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2601 log only when "message" or "log_message" are set. However, if one of the
2602 conditions was an address verification, or a call to a nested ACL, the
2603 messages generated by the underlying test were being passed through. This
2604 no longer happens. The underlying message is available in $acl_verify_
2605 message for both "message" and "log_message" expansions, so it can be
2606 passed through if needed.
2608 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2609 new expansion $bh_ to give the encoded byte string without charset
2610 translation. Translation happens only if iconv() is available; HAVE_ICONV
2611 indicates this at build time. HEADERS_CHARSET gives the charset to
2612 translate to; headers_charset can change it in the configuration, and
2613 "headers charset" can change it in an individual filter file.
2615 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2616 that creates RFC 2047 encoded "words" labels them as that charset instead
2617 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2618 expansion operator; (ii) when Exim creates a From: line for a local
2619 message; (iii) when a header line is rewritten to include a "phrase" part.
2621 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2622 buggy, causing it to skip the first lines of messages whose message ID
2623 ended in 'D'. This would not have bitten before Exim release 4.14, because
2624 message IDs were unlikely to end in 'D' before then. The effect was to have
2625 incorrect size information for certain domains.
2627 45. #include "config.h" was missing at the start of the crypt16.c module. This
2628 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2631 46. If there was a timeout during a "random" callout check, Exim treated it as
2632 a failure of the random address, and carried on sending RSET and the real
2633 address. If the delay was just some slowness somewhere, the response to the
2634 original RCPT would be taken as a response to RSET and so on, causing
2635 mayhem of various kinds.
2637 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2638 when I implemented it. It didn't allow for the fact that some option values
2639 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2640 the right test for positive values.
2642 48. Domain names in DNS records are case-independent. Exim always looks them up
2643 in lower case. Some resolvers return domain names in exactly the case they
2644 appear in the zone file, that is, they may contain uppercase letters. Not
2645 all resolvers do this - some return always lower case. Exim was treating a
2646 change of case by a resolver as a change of domain, similar to a widening
2647 of a domain abbreviation. This triggered its re-routing code and so it was
2648 trying to route what was effectively the same domain again. This normally
2649 caused routing to fail (because the router wouldn't handle the domain
2650 twice). Now Exim checks for this case specially, and just changes the
2651 casing of the domain that it ultimately uses when it transmits the message
2654 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2657 50. If a filter generated a file delivery with a non-absolute name (possible if
2658 no home directory exists for the router), the forbid_file option was not
2661 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2662 the existing equality matching.
2664 52. Exim was using ints instead of ino_t variables in some places where it was
2665 dealing with inode numbers.
2667 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2668 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2669 called TMPDIR, and if it finds it is different, it changes its value.
2671 54. The smtp_printf() function is now made available to local_scan() so
2672 additional output lines can be written before returning. There is also an
2673 smtp_fflush() function to enable the detection of a dropped connection.
2674 The variables smtp_input and smtp_batched_input are exported to
2677 55. Changed the default runtime configuration: the message "Unknown user"
2678 has been removed from the ACL, and instead placed on the localuser router,
2679 using the cannot_route_message feature. This means that any verification
2680 failures that generate their own messages won't get overridden. Similarly,
2681 the "Unrouteable address" message that was in the ACL for unverifiable
2682 relay addresses has also been removed.
2684 56. Added hosts_avoid_esmtp to the smtp transport.
2686 57. The exicyclog script was not checking for the esoteric option
2687 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2688 will work only if exicyclog is run under the appropriate euid.
2690 58. Following a discussion on the list, the rules by which Exim recognises line
2691 endings on incoming messages have been changed. The -dropcr and drop_cr
2692 options are now no-ops, retained only for backwards compatibility. The
2693 following line terminators are recognized: LF CRLF CR. However, special
2694 processing applies to CR:
2696 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2697 nor a local message in the state where . is a terminator.
2699 (ii) If a bare CR is encountered in a header line, an extra space is added
2700 after the line terminator so as not to end the header. The reasoning
2701 behind this is that bare CRs in header lines are most likely either
2702 to be mistakes, or people trying to play silly games.
2704 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2705 was being incorrectly given as 18 bytes larger than it should have been.
2706 This is a VOB (very old bug).
2708 60. This may never have affected anything current, but just in case it has:
2709 When the local host is found other than at the start of a list of hosts,
2710 the local host, those with the same MX, and any that follow, are discarded.
2711 When the list in question was part of a longer list of hosts, the following
2712 hosts (not currently being processed) were also being discarded. This no
2713 longer happens. I'm not sure if this situation could ever has previously
2716 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2719 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2720 header. This is something that is recommended in a new Internet Draft, and
2721 is something that is documented as being done by Sendmail. There are two
2722 possible values. For messages generated by the autoreply transport, Exim
2725 Auto-Submitted: auto-replied
2727 whereas for all other generated messages (e.g. bounces) it adds
2729 Auto-Submitted: auto-generated
2731 63. The "personal" condition in filters now includes a test for the
2732 Auto-Submitted: header. If it contains the string "auto-" the message it
2733 not considered personal.
2735 64. Added rcpt_include_affixes as a generic transport option.
2737 65. Added queue_only_override (default true).
2739 66. Added the syslog_duplication option.
2741 67. If what should have been the first header line of a message consisted of
2742 a space followed by a colon, Exim was mis-interpreting it as a header line.
2743 It isn't of course - it is syntactically invalid and should therefore be
2744 treated as the start of the message body. The misbehaviour could have
2745 caused a number of strange effects, including loss of data in subsequent
2746 header lines, and spool format errors.
2748 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2749 client host had authenticated. This control can now be exercised by an ACL
2750 for more flexibility.
2752 69. By default, callouts do not happen when testing with -bh. There is now a
2753 variant, -bhc, which does actually run the callout code, including
2754 consulting and updating the callout cache.
2756 70. Added support for saslauthd authentication, courtesy of Alexander
2759 71. If statvfs() failed on the spool or log directories while checking their
2760 size for availability, Exim confusingly gave the error "space shortage".
2761 Furthermore, in debugging mode it crashed with a floating point exception.
2762 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2763 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2764 serious problem, Exim now writes to the main and panic logs when this
2765 happens, with details of the failure. It then refuses to accept the
2766 incoming message, giving the message "spool directory problem" or "log
2767 directory problem" with a 421 code for SMTP messages.
2769 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2770 0, 1, and 2 exist, because some OS complain for execs without them (see
2771 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2772 descriptors. However, the code omitted to check that the open succeeded,
2773 causing mysterious errors if for some reason the permissions on /dev/null
2774 got screwed. Now Exim writes a message to the main and panic logs, and
2775 bombs out if it can't open /dev/null.
2777 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2778 interact so that it is all more flexible. It is supposed to remain
2779 backwards compatible. Also added extra_local_interfaces.
2781 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2782 to bomb out with an assertion failure - to the client this appears as a
2783 connection drop. This problem occurs in the part of the code that was taken
2784 from the Samba project. Fortunately, the assertion is in a very simple
2785 function, so I have fixed this by reproducing the function inline in the
2786 one place where it is called, and arranging for authentication to fail
2787 instead of killing the process with assert().
2789 75. The SPA client code was not working when the server requested OEM rather
2790 than Unicode encoding.
2792 76. Added code to make require_files with a specific uid setting more usable in
2793 the case where statting the file as root fails - usually a non-root-mounted
2794 NFS file system. When this happens and the failure is EACCES, Exim now
2795 forks a subprocess and does the per-uid checking as the relevant uid.
2797 77. Added process_log_path.
2799 78. If log_file_path was not explicitly set, a setting of check_log_space or
2800 check_log_inodes was ignored.
2802 79. If a space check for the spool or log partitions fails, the incident is now
2803 logged. Of course, in the latter case the data may get lost...
2805 80. Added the %p formatting code to string_format() so that it can be used to
2806 print addresses in debug_print(). Adjusted all the address printing in the
2807 debugging in store.c to use %p rather than %d.
2809 81. There was a concern that a line of code in smtp_in.c could overflow a
2810 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2811 initially expressed, the concern was not well-founded, because trailing
2812 spaces are removed early. However, if the trailing spaces were followed by
2813 a NULL, they did not get removed, so the overflow was possible. Two fixes
2816 (a) I re-wrote the offending code in a cleaner fashion.
2817 (b) If an incoming SMTP command contains a NULL character, it is rejected
2820 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2821 runs initgroups(), so that if the Exim user is in any additional groups,
2822 they will be used during message reception.
2828 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob