1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.141 2005/05/23 16:58:55 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
12 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
13 command itself ended in error, the underlying message about the transport
14 filter timeout was being overwritten with the pipe command error. Now the
15 underlying error message should be appended to the second error message.
17 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
19 PH/02 Reduce the amount of output that "make" produces by default. Full output
20 can still be requested.
22 PH/03 The warning log line about a condition test deferring for a "warn" verb
23 was being output only once per connection, rather than after each
24 occurrence (because it was using the same function as for successful
25 "warn" verbs). This seems wrong, so I have changed it.
27 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
28 it should not have, which might have caused a crash in the right
29 circumstances, but probably never did.
31 PH/04 Installed a modified version of Tony Finch's patch to make submission
32 mode fix the return path as well as the Sender: header line, and to
33 add a /name= option so that you can make the user's friendly name appear
36 TF/03 Added the control = fakedefer ACL modifier.
38 TF/04 Added the ratelimit ACL condition. See NewStuff for details.
44 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
45 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
47 TK/02 Fix ACL "control" statement not being available in MIME ACL.
49 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
51 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
52 to test Sieve filters that use "vacation".
54 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
55 that changes the way the GnuTLS parameters are stored in the cache file.
56 The new format can be generated externally. For backward compatibility,
57 if the data in the cache doesn't make sense, Exim assumes it has read an
58 old-format file, and it generates new data and writes a new file. This
59 means that you can't go back to an older release without removing the
62 PH/03 A redirect router that has both "unseen" and "one_time" set does not
63 work if there are any delivery delays because "one_time" forces the
64 parent to be marked "delivered", so its unseen clone is never tried
65 again. For this reason, Exim now forbids the simultaneous setting of
68 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
69 redirected to themselves ("homonym" addresses). Read the long ChangeLog
70 entry if you want to know the details. The fix, however, neglected to
71 consider the case when local delivery batching is involved. The test for
72 "previously delivered" was not happening when checking to see if an
73 address could be batched with a previous (undelivered) one; under
74 certain circumstances this could lead to multiple deliveries to the same
77 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
78 in its include files, and this causes problems building Exim.
80 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
81 header_syntax) but Exim was just ignoring anything given after a slash.
82 In particular, this caused confusion with an attempt to use "verify =
83 reverse_host_lookup/defer_ok". An error is now given when options are
84 supplied for verify items that do not have them. (Maybe reverse_host_
85 lookup should have a defer_ok option, but that's a different point.)
87 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
88 defined by RFC 821) to 2048, because there were problems with some AUTH
89 commands, and RFC 1869 says the size should be increased for extended
90 SMTP commands that take arguments.
92 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
95 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
96 "unknown" error; now it says that the functionality isn't in the binary.
98 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
99 an address' error message when a string expansion fails (syntax or
100 whatever). Otherwise the password may appear in the log. Following change
101 PH/42 below, there is no longer a chance of it appearing in a bounce
104 PH/11 Installed exipick version 20050225.0 from John Jetmore.
106 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
107 of its addresses was ever tried. (Bugzilla bug #2.)
109 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
110 the result incorrectly in the debug output. (It correctly added a newline
111 to what was transported.)
113 TF/01 Added $received_time.
115 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
116 commented out examples of how to interface to a virus scanner and to
117 SpamAssassin. Also added commented examples of av_scanner and
118 spamd_address settings.
120 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
121 and controls are allowed in which ACLs. There were a couple of minor
122 errors. Some of the entries in the conditions table (which is a table of
123 where they are NOT allowed) were getting very unwieldy; rewrote them as a
124 negation of where the condition IS allowed.
126 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
128 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
129 header file does not have a version number, so I've had to invent a new
130 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
131 API. The code is untested by me (my Linux distribution still has 0.3.2 of
132 radiusclient), but it was contributed by a Radius user.
134 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
135 files or directories, for OpenSSL.
137 PH/19 When an Exim process that is running as root has to create an Exim log
138 file, it does so in a subprocess that runs as exim:exim so as to get the
139 ownership right at creation (otherwise, other Exim processes might see
140 the file with the wrong ownership). There was no test for failure of this
141 fork() call, which would lead to the process getting stuck as it waited
142 for a non-existent subprocess. Forks do occasionally fail when resources
143 run out. I reviewed all the other calls to fork(); they all seem to check
146 PH/20 When checking for unexpected SMTP input at connect time (before writing
147 the banner), Exim was not dealing correctly with a non-positive return
148 from the read() function. If the client had disconnected by this time,
149 the result was a log entry for a synchronization error with an empty
150 string after "input=" when read() returned zero. If read() returned -1
151 (an event I could not check), uninitialized data bytes were printed.
152 There were reports of junk text (parts of files, etc) appearing after
155 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
157 PH/22 Added support for macro redefinition, and (re)definition in between
158 driver and ACL definitions.
160 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
161 forgetting to use the resulting value; it was using the unexpanded value.
163 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
164 hadn't been configured. The fix is from Juergen Kreileder, who
165 understands it better than I do:
167 "Here's what I see happening with three configured cyrus_sasl
168 authenticators configured (plain, login, cram-md5):
170 On startup auth_cyrus_sasl_init() gets called for each of these.
171 This means three calls to sasl_listmech() without a specified mech_list.
172 => SASL tests which mechs of all available mechs actually work
173 => three warnings about OTP not working
174 => the returned list contains: plain, login, cram-md5, digest-md5, ...
176 With the patch, sasl_listmech() also gets called three times. But now
177 SASL's mech_list option is set to the server_mech specified in the the
178 authenticator. Or in other words, the answer from sasl_listmech()
179 gets limited to just the mech you're testing for (which is different
181 => the return list contains just 'plain' or 'login', 'cram-md5' or
182 nothing depending on the value of ob->server_mech.
184 I've just tested the patch: Authentication still works fine,
185 unavailable mechs specified in the exim configuration are still
186 caught, and the auth.log warnings about OTP are gone."
188 PH/25 When debugging is enabled, the contents of the command line are added
189 to the debugging output, even when log_selector=+arguments is not
192 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
193 answer is "GNU", and only if the return is "GNU/something" is the answer
196 PH/27 $acl_verify_message is now set immediately after the failure of a
197 verification in an ACL, and so is available in subsequent modifiers. In
198 particular, the message can be preserved by coding like this:
200 warn !verify = sender
201 set acl_m0 = $acl_verify_message
203 Previously, $acl_verify_message was set only while expanding "message"
204 and "log_message" when a very denied access.
206 PH/28 Modified OS/os.c-Linux with
208 -#ifndef OS_LOAD_AVERAGE
209 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
211 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
212 nomenclature these days.)
214 PH/29 Installed patch from the Sieve maintainer that adds the options
215 sieve_useraddress and sieve_subaddress to the redirect router.
217 PH/30 In these circumstances:
218 . Two addresses routed to the same list of hosts;
219 . First host does not offer TLS;
220 . First host accepts first address;
221 . First host gives temporary error to second address;
222 . Second host offers TLS and a TLS session is established;
223 . Second host accepts second address.
224 Exim incorrectly logged both deliveries with the TLS parameters (cipher
225 and peerdn, if requested) that were in fact used only for the second
228 PH/31 When doing a callout as part of verifying an address, Exim was not paying
229 attention to any local part prefix or suffix that was matched by the
230 router that accepted the address. It now behaves in the same way as it
231 does for delivery: the affixes are removed from the local part unless
232 rcpt_include_affixes is set on the transport.
234 PH/32 Add the sender address, as F=<...>, to the log line when logging a
235 timeout during the DATA phase of an incoming message.
237 PH/33 Sieve envelope tests were broken for match types other than :is. I have
238 applied a patch sanctioned by the Sieve maintainer.
240 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
241 the uid or gid is negative. A case of a negative gid caused this to be
242 noticed. The fix allows for either to be negative.
244 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
245 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
248 PH/36 The change PH/12 above was broken. Fixed it.
250 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
251 the grounds that routing the same address twice would always produce the
252 same answer. This might have been true once, but it is certainly no
253 longer true now. Routing a child address may depend on the previous
254 routing that produced that child. Some complicated redirection strategies
255 went wrong when messages had multiple recipients, and made Exim's
256 behaviour dependent on the order in which the addresses were given.
258 I have moved the duplicate checking until after the routing is complete.
259 Exim scans the addresses that are assigned to local and remote
260 transports, and removes any duplicates. This means that more work will be
261 done, as duplicates will always all be routed, but duplicates are
262 presumably rare, so I don't expect this is of any significance.
264 For deliveries to pipes, files, and autoreplies, the duplicate checking
265 still happens during the routing process, since they are not going to be
268 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
269 It corrects a timeout issue with spamd. This is Ian's comment: "The
270 background is that sometimes spamd either never reads data from a
271 connection it has accepted, or it never writes response data. The exiscan
272 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
273 blindly assumes that writes won't block so it may never time out."
275 PH/39 Allow G after quota size as well as K and M.
277 PH/40 The value set for $authenticated_id in an authenticator may not contain
278 binary zeroes or newlines because the value is written to log lines and
279 to spool files. There was no check on this. Now the value is run through
280 the string_printing() function so that such characters are converted to
281 printable escape sequences.
283 PH/41 $message_linecount is a new variable that contains the total number of
284 lines in the message. Compare $body_linecount, which is the count for the
287 PH/42 Exim no longer gives details of delivery errors for specific addresses in
288 bounce and delay warning messages, except in certain special cases, which
291 (a) An SMTP error message from a remote host;
292 (b) A message specified in a :fail: redirection;
293 (c) A message specified in a "fail" command in a system filter;
294 (d) A message specified in a FAIL return from the queryprogram router;
295 (e) A message specified by the cannot_route_message router option.
297 In these cases only, Exim does include the error details in bounce and
298 warning messages. There are also a few cases where bland messages such
299 as "unrouteable address" or "local delivery error" are given.
301 PH/43 $value is now also set for the "else" part of a ${run expansion.
303 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
304 being worked on, but at least Exim now implements the latest version to
307 PH/45 In a pipe transport, although a timeout while waiting for the pipe
308 process to complete was treated as a delivery failure, a timeout while
309 writing the message to the pipe was logged, but erroneously treated as a
310 successful delivery. Such timeouts include transport filter timeouts. For
311 consistency with the overall process timeout, these timeouts are now
312 treated as errors, giving rise to delivery failures by default. However,
313 there is now a new Boolean option for the pipe transport called
314 timeout_defer, which, if set TRUE, converts the failures into defers for
315 both kinds of timeout. A transport filter timeout is now identified in
318 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
319 systems where "make" and "gmake" are different, calling "gmake" at top
320 level broke things. I've arranged for the value of $(MAKE) to be passed
321 from the Makefile to this script so that it can call the same version of
325 A note about Exim versions 4.44 and 4.50
326 ----------------------------------------
328 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
329 changes of various kinds. As a consequence, a big documentation update was
330 needed. This delayed the release for rather longer than seemed good, especially
331 in the light of a couple of (minor) security issues. Therefore, the changes
332 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
333 release. So 4.44 and 4.50 are in effect two different branches that both start
336 I have left the 4.50 change log unchanged; it contains all the changes since
337 4.43. The change log for 4.44 is below; many of its items are identical to
338 those for 4.50. This seems to be the most sensible way to preserve the
339 historical information.
345 1. Minor wording change to the doc/README.SIEVE file.
347 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
348 computation of the current number of files was incorrect.
350 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
351 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
354 4. Give more explanation in the error message when the command for a transport
355 filter fails to execute.
357 5. There are several places where Exim runs a non-Exim command in a
358 subprocess. The SIGUSR1 signal should be disabled for these processes. This
359 was being done only for the command run by the queryprogram router. It is
360 now done for all such subprocesses. The other cases are: ${run, transport
361 filters, and the commands run by the lmtp and pipe transports.
363 6. Added CONFIGURE_GROUP build-time option.
365 7. Some older OS have a limit of 256 on the maximum number of file
366 descriptors. Exim was using setrlimit() to set 1000 as a large value
367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
368 systems. I've change it so that if it can't get 1000, it tries for 256.
370 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
371 was an oversight, and furthermore, ever since the addition of extra
372 controls (e.g. 4.43/32), the checks on when to allow different forms of
373 "control" were broken. There should now be diagnostics for all cases when a
374 control that does not make sense is encountered.
376 9. Added the /retain_sender option to "control=submission".
378 10. $recipients is now available in the predata ACL (oversight).
380 11. Tidy the search cache before the fork to do a delivery from a message
381 received from the command line. Otherwise the child will trigger a lookup
382 failure and thereby defer the delivery if it tries to use (for example) a
383 cached ldap connection that the parent has called unbind on.
385 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
386 of $address_data from the recipient verification was clobbered by the
389 13. The value of address_data from a sender verification is now available in
390 $sender_address_data in subsequent conditions in the ACL statement.
392 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
394 15. Added a new option "connect=<time>" to callout options, to set a different
397 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
398 was its contents. (It was OK if the option was not defined at all.)
400 17. A "Completed" log line is now written for messages that are removed from
401 the spool by the -Mrm option.
403 18. New variables $sender_verify_failure and $recipient_verify_failure contain
404 information about exactly what failed.
406 19. Added -dd to debug only the daemon process.
408 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
409 handles timeouts, both on the server side and network timeouts. Renamed the
410 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
412 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
413 It is now set to "smtps".
415 22. $host_address is now set to the target address during the checking of
418 23. When checking ignore_target_hosts for an ipliteral router, no host name was
419 being passed; this would have caused $sender_host_name to have been used if
420 matching the list had actually called for a host name (not very likely,
421 since this list is usually IP addresses). A host name is now passed as
424 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
425 code that specifies a non-restarting handler (typically sigaction() in
426 modern systems) in an attempt to fix a rare and obscure crash bug.
428 25. Narrowed the window for a race in the daemon that could cause it to ignore
429 SIGCHLD signals. This is not a major problem, because they are used only to
430 wake it up if nothing else does.
432 26. A malformed maildirsize file could cause Exim to calculate negative values
433 for the mailbox size or file count. Odd effects could occur as a result.
434 The maildirsize information is now recalculated if the size or filecount
437 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
438 support for a long time. Removed HAVE_SYS_VFS_H.
440 28. Installed the latest version of exipick from John Jetmore.
442 29. In an address list, if the pattern was not a regular expression, an empty
443 subject address (from a bounce message) matched only if the pattern was an
444 empty string. Non-empty patterns were not even tested. This was the wrong
445 because it is perfectly reasonable to use an empty address as part of a
446 database query. An empty address is now tested by patterns that are
447 lookups. However, all the other forms of pattern expect the subject to
448 contain a local part and a domain, and therefore, for them, an empty
449 address still always fails if the pattern is not itself empty.
451 30. Exim went into a mad DNS loop when attempting to do a callout where the
452 host was specified on an smtp transport, and looking it up yielded more
455 31. Re-factored the code for checking spool and log partition space into a
456 function that finds that data and another that does the check. The former
457 is then used to implement four new variables: $spool_space, $log_space,
458 $spool_inodes, and $log_inodes.
460 32. The RFC2047 encoding function was originally intended for short strings
461 such as real names; it was not keeping to the 75-character limit for
462 encoded words that the RFC imposes. It now respects the limit, and
463 generates multiple encoded words if necessary. To be on the safe side, I
464 have increased the buffer size for the ${rfc2047: expansion operator from
467 33. It is now permitted to omit both strings after an "if" condition; if the
468 condition is true, the result is "true". As before, when the second string
469 is omitted, a false condition yields an empty string. This makes it less
470 cumbersome to write custom ACL and router conditions.
472 34. Failure to deliver a bounce message always caused it to be frozen, even if
473 there was an errors_to setting on the router. The errors_to setting is now
476 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
477 canonical form (fully expanded) before being placed in
478 $sender_host_address.
480 36. The table in the code that translates DNS record types into text (T_A to
481 "A" for instance) was missing entries for NS and CNAME. It is just possible
482 that this could have caused confusion if both these types were looked up
483 for the same domain, because the text type is used as part of Exim's
484 per-process caching. But the chance of anyone hitting this buglet seems
487 37. The dnsdb lookup has been extended in a number of ways.
489 (1) There is a new type, "zns", which walks up the domain tree until it
490 finds some nameserver records. It should be used with care.
492 (2) There is a new type, "mxh", which is like "mx" except that it returns
493 just the host names, not the priorities.
495 (3) It is now possible to give a list of domains (or IP addresses) to be
496 looked up. The behaviour when one of the lookups defers can be
497 controlled by a keyword.
499 (4) It is now possible to specify the separator character for use when
500 multiple records are returned.
502 38. The dnslists ACL condition has been extended: it is now possible to supply
503 a list of IP addresses and/or domains to be looked up in a particular DNS
506 39. Added log_selector=+queue_time_overall.
508 40. When running the queue in the test harness, wait just a tad after forking a
509 delivery process, to get repeatability of debugging output.
511 41. Include certificate and key file names in error message when GnuTLS fails
512 to set them up, because the GnuTLS error message doesn't include the name
513 of the failing file when there is a problem reading it.
515 42. Allow both -bf and -bF in the same test run.
517 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
519 44. The "Exiscan patch" is now merged into the mainline Exim source.
521 45. Sometimes the final signoff response after QUIT could fail to get
522 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
523 < 0 before doing a fflush(). This bug looks as though it goes back to the
524 introduction of TLS in release 3.20, but "sometimes" must have been rare
525 because the tests only now provoked it.
527 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
528 (this can affect the format of dates).
530 47. exim_tidydb, when checking for the continued existence of a message for
531 which it has found a message-specific retry record, was not finding
532 messages that were in split spool directories. Consequently, it was
533 deleting retry records that should have stayed in existence.
535 48. Steve fixed some bugs in eximstats.
537 49. The SPA authentication driver was not abandoning authentication and moving
538 on to the next authenticator when an expansion was forced to fail,
539 contradicting the general specification for all authenticators. Instead it
540 was generating a temporary error. It now behaves as specified.
542 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
543 (the order specifies the preference for clients). The order is now AES256,
544 AES128, 3DES, ARCFOUR128.
546 51. Small patch to Sieve code - explicitly set From: when generating an
549 52. Exim crashed if a remote delivery caused a very long error message to be
550 recorded - for instance if somebody sent an entire SpamAssassin report back
551 as a large number of 550 error lines. This bug was coincidentally fixed by
552 increasing the size of one of Exim's internal buffers (big_buffer) that
553 happened as part of the Exiscan merge. However, to be on the safe side, I
554 have made the code more robust (and fixed the comments that describe what
557 53. Now that there can be additional text after "Completed" in log lines (if
558 the queue_time_overall log selector is set), a one-byte patch to exigrep
559 was needed to allow it to recognize "Completed" as not the last thing in
562 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
563 patch that reportedly fixes this has been added. I am not expert enough to
564 create a test for it. This is what the patch creator wrote:
566 "I found a little strange behaviour of ldap code when working with
567 Windows 2003 AD Domain, where users was placed in more than one
568 Organization Units. When I tried to give exim partial DN, the exit code
569 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
570 But simultaneously result of request was absolutely normal ldap result,
571 so I produce this patch..."
573 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
574 REFERENCE, so I have modified the code to exclude the patch when that macro
577 55. Some experimental protocols are using DNS PTR records for new purposes. The
578 keys for these records are domain names, not reversed IP addresses. The
579 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
580 leaves it alone. Component reversal etc. now happens only for IP addresses.
583 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
585 57. Double the size of the debug message buffer (to 2048) so that more of very
586 long debug lines gets shown.
588 58. The exicyclog utility now does better if the number of log files to keep
589 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
591 59. Two changes related to the smtp_active_hostname option:
593 (1) $smtp_active_hostname is now available as a variable.
594 (2) The default for smtp_banner uses $smtp_active_hostname instead
595 of $primary_hostname.
597 60. The host_aton() function is supposed to be passed a string that is known
598 to be a valid IP address. However, in the case of IPv6 addresses, it was
599 not checking this. This is a hostage to fortune. Exim now panics and dies
600 if the condition is not met. A case was found where this could be provoked
601 from a dnsdb PTR lookup with an IPv6 address that had more than 8
602 components; fortuitously, this particular loophole had already been fixed
603 by change 4.50/55 above.
605 If there are any other similar loopholes, the new check in host_aton()
606 itself should stop them being exploited. The report I received stated that
607 data on the command line could provoke the exploit when Exim was running as
608 exim, but did not say which command line option was involved. All I could
609 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
613 61. There was a buffer overflow vulnerability in the SPA authentication code
614 (which came originally from the Samba project). I have added a test to the
615 spa_base64_to_bits() function which I hope fixes it.
618 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
619 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
621 63. The daemon start-up calls getloadavg() while still root for those OS that
622 need the first call to be done as root, but it missed one case: when
623 deliver_queue_load_max is set with deliver_drop_privilege. This is
624 necessary for the benefit of the queue runner, because there is no re-exec
625 when deliver_drop_privilege is set.
627 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
630 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
631 was in use, was not putting the data itself into the right store pool;
632 consequently, it could be overwritten for a subsequent message in the same
633 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
636 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
638 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
639 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
640 the function were treating the return as a boolean value, which happened to
641 work because 0=false and not-0=true, but is not correct code.
643 68. The host_aton() function was not handling scoped IPv6 addresses (those
644 with, for example, "%eth0" on the end) correctly.
646 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
647 negated items (that is, ~something) in unsigned ints. Some compilers
648 apparently mutter when there is no cast.
650 70. If an address verification called from an ACL failed, and did not produce a
651 user-specific message (i.e. there was only a "system" message), nothing was
652 put in $acl_verify_message. In this situation, it now puts the system
655 71. Change 4.23/11 added synchronization checking at the start of an SMTP
656 session; change 4.31/43 added the unwanted input to the log line - except
657 that it did not do this in the start of session case. It now does.
659 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
660 This is wrong and can cause the other end to generate a synchronization
661 error if it is another Exim or anything else that does the synchronization
662 check. A QUIT command is no longer sent after a timeout.
664 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
667 74. The defer_ok option of callout verification was not working if it was used
668 when verifying addresses in header lines, that is, for this case:
670 verify = header_sender/callout=defer_ok
672 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
673 those file descriptors could be used for SMTP connections. If anything
674 wrote to stderr (the example that came up was "warn" in embedded Perl), it
675 could be sent to the SMTP client, causing chaos. The daemon now opens
676 stdin, stdout, and stderr to /dev/null when it puts itself into the
679 76. Arrange for output from Perl's "warn" command to be written to Exim's main
680 log by default. The user can override this with suitable Perl magic.
682 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
683 the log message when discard triggers, was not working for the DATA ACL or
684 for the non-SMTP ACL.
686 78. Error message wording change in sieve.c.
688 79. If smtp_accept_max_per_host was set, the number of connections could be
689 restricted to fewer than expected, because the daemon was trying to set up
690 a new connection before checking whether the processes handling previous
691 connections had finished. The check for completed processes is now done
692 earlier. On busy systems, this bug wouldn't be noticed because something
693 else would have woken the daemon, and it would have reaped the completed
696 80. If a message was submitted locally by a user whose login name contained one
697 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
698 It caused a spool format error. I have fixed the spool reading code. A
699 related problem was that the "from" clause in the Received: line became
700 illegal because of the space(s). It is now covered by ${quote_local_part.
702 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
705 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
707 83. Patch from Sieve maintainer to fix unterminated string problem in
710 84. Some minor changes to the Linux configuration files to help with other
711 OS variants using glibc.
713 85. One more patch for Sieve to update vacation handling to latest spec.
716 ----------------------------------------------------
717 See the note above about the 4.44 and 4.50 releases.
718 ----------------------------------------------------
724 1. Change 4.43/35 introduced a bug that caused file counts to be
725 incorrectly computed when quota_filecount was set in an appendfile
728 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
729 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
732 3. Give more explanation in the error message when the command for a transport
733 filter fails to execute.
735 4. There are several places where Exim runs a non-Exim command in a
736 subprocess. The SIGUSR1 signal should be disabled for these processes. This
737 was being done only for the command run by the queryprogram router. It is
738 now done for all such subprocesses. The other cases are: ${run, transport
739 filters, and the commands run by the lmtp and pipe transports.
741 5. Some older OS have a limit of 256 on the maximum number of file
742 descriptors. Exim was using setrlimit() to set 1000 as a large value
743 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
744 systems. I've change it so that if it can't get 1000, it tries for 256.
746 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
747 was an oversight, and furthermore, ever since the addition of extra
748 controls (e.g. 4.43/32), the checks on when to allow different forms of
749 "control" were broken. There should now be diagnostics for all cases when a
750 control that does not make sense is encountered.
752 7. $recipients is now available in the predata ACL (oversight).
754 8. Tidy the search cache before the fork to do a delivery from a message
755 received from the command line. Otherwise the child will trigger a lookup
756 failure and thereby defer the delivery if it tries to use (for example) a
757 cached ldap connection that the parent has called unbind on.
759 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
760 of $address_data from the recipient verification was clobbered by the
763 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
764 was its contents. (It was OK if the option was not defined at all.)
766 11. A "Completed" log line is now written for messages that are removed from
767 the spool by the -Mrm option.
769 12. $host_address is now set to the target address during the checking of
772 13. When checking ignore_target_hosts for an ipliteral router, no host name was
773 being passed; this would have caused $sender_host_name to have been used if
774 matching the list had actually called for a host name (not very likely,
775 since this list is usually IP addresses). A host name is now passed as
778 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
779 code that specifies a non-restarting handler (typically sigaction() in
780 modern systems) in an attempt to fix a rare and obscure crash bug.
782 15. Narrowed the window for a race in the daemon that could cause it to ignore
783 SIGCHLD signals. This is not a major problem, because they are used only to
784 wake it up if nothing else does.
786 16. A malformed maildirsize file could cause Exim to calculate negative values
787 for the mailbox size or file count. Odd effects could occur as a result.
788 The maildirsize information is now recalculated if the size or filecount
791 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
792 support for a long time. Removed HAVE_SYS_VFS_H.
794 18. Updated exipick to current release from John Jetmore.
796 19. Allow an empty sender to be matched against a lookup in an address list.
797 Previously the only cases considered were a regular expression, or an
800 20. Exim went into a mad DNS lookup loop when doing a callout where the
801 host was specified on the transport, if the DNS lookup yielded more than
804 21. The RFC2047 encoding function was originally intended for short strings
805 such as real names; it was not keeping to the 75-character limit for
806 encoded words that the RFC imposes. It now respects the limit, and
807 generates multiple encoded words if necessary. To be on the safe side, I
808 have increased the buffer size for the ${rfc2047: expansion operator from
811 22. Failure to deliver a bounce message always caused it to be frozen, even if
812 there was an errors_to setting on the router. The errors_to setting is now
815 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
816 canonical form (fully expanded) before being placed in
817 $sender_host_address.
819 24. Updated eximstats to version 1.33
821 25. Include certificate and key file names in error message when GnuTLS fails
822 to set them up, because the GnuTLS error message doesn't include the name
823 of the failing file when there is a problem reading it.
825 26. Expand error message when OpenSSL has problems setting up cert/key files.
828 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
829 (this can affect the format of dates).
831 28. exim_tidydb, when checking for the continued existence of a message for
832 which it has found a message-specific retry record, was not finding
833 messages that were in split spool directories. Consequently, it was
834 deleting retry records that should have stayed in existence.
836 29. eximstats updated to version 1.35
837 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
838 1.35 - bugfix such that pie charts by volume are generated correctly
840 30. The SPA authentication driver was not abandoning authentication and moving
841 on to the next authenticator when an expansion was forced to fail,
842 contradicting the general specification for all authenticators. Instead it
843 was generating a temporary error. It now behaves as specified.
845 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
846 (the order specifies the preference for clients). The order is now AES256,
847 AES128, 3DES, ARCFOUR128.
849 31. Small patch to Sieve code - explicitly set From: when generating an
852 32. Exim crashed if a remote delivery caused a very long error message to be
853 recorded - for instance if somebody sent an entire SpamAssassin report back
854 as a large number of 550 error lines. This bug was coincidentally fixed by
855 increasing the size of one of Exim's internal buffers (big_buffer) that
856 happened as part of the Exiscan merge. However, to be on the safe side, I
857 have made the code more robust (and fixed the comments that describe what
860 33. Some experimental protocols are using DNS PTR records for new purposes. The
861 keys for these records are domain names, not reversed IP addresses. The
862 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
863 leaves it alone. Component reversal etc. now happens only for IP addresses.
866 34. The host_aton() function is supposed to be passed a string that is known
867 to be a valid IP address. However, in the case of IPv6 addresses, it was
868 not checking this. This is a hostage to fortune. Exim now panics and dies
869 if the condition is not met. A case was found where this could be provoked
870 from a dnsdb PTR lookup with an IPv6 address that had more than 8
871 components; fortuitously, this particular loophole had already been fixed
872 by change 4.50/55 or 4.44/33 above.
874 If there are any other similar loopholes, the new check in host_aton()
875 itself should stop them being exploited. The report I received stated that
876 data on the command line could provoke the exploit when Exim was running as
877 exim, but did not say which command line option was involved. All I could
878 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
882 35. There was a buffer overflow vulnerability in the SPA authentication code
883 (which came originally from the Samba project). I have added a test to the
884 spa_base64_to_bits() function which I hope fixes it.
887 36. The daemon start-up calls getloadavg() while still root for those OS that
888 need the first call to be done as root, but it missed one case: when
889 deliver_queue_load_max is set with deliver_drop_privilege. This is
890 necessary for the benefit of the queue runner, because there is no re-exec
891 when deliver_drop_privilege is set.
893 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
894 was in use, was not putting the data itself into the right store pool;
895 consequently, it could be overwritten for a subsequent message in the same
896 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
899 38. Sometimes the final signoff response after QUIT could fail to get
900 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
901 < 0 before doing a fflush(). This bug looks as though it goes back to the
902 introduction of TLS in release 3.20, but "sometimes" must have been rare
903 because the tests only now provoked it.
909 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
910 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
911 Now it returns an integer. A number of calls were still expecting a T/F
912 return. Fortuitously, in all cases, the tests worked in OK situations,
913 which is the norm. However, things would have gone wrong on any write
914 failures on the smtp file descriptor. This function is used when sending
915 messages over SMTP and also when doing verify callouts.
917 2. When Exim is called to do synchronous delivery of a locally submitted
918 message (the -odf or -odi options), it no longer closes stderr before doing
921 3. Implemented the mua_wrapper option.
923 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
925 5. Implemented the functions header_remove(), header_testname(),
926 header_add_at_position(), and receive_remove_recipient(), and exported them
929 6. If an ACL "warn" statement specified the addition of headers, Exim already
930 inserted X-ACL-Warn: at the start if there was no header name. However, it
931 was not making this test for the second and subsequent header lines if
932 there were newlines in the string. This meant that an invalid header could
933 be inserted if Exim was badly configured.
935 7. Allow an ACL "warn" statement to add header lines at the start or after all
936 the Received: headers, as well as at the end.
938 8. Added the rcpt_4xx retry error code.
940 9. Added postmaster_mailfrom=xxx to callout verification option.
942 10. Added mailfrom=xxxx to the callout verification option, for verify=
945 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
946 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
948 12. Inserted some casts to stop certain compilers warning when using pointer
949 differences as field lengths or precisions in printf-type calls (mostly
950 affecting debugging statements).
952 13. Added optional readline() support for -be (dynamically loaded).
954 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
955 same clock tick as a message's arrival, so that its received time was the
956 same as the "first fail" time on the retry record, and that message
957 remained on the queue past the ultimate address timeout, every queue runner
958 would try a delivery (because it was past the ultimate address timeout) but
959 after another failure, the ultimate address timeout, which should have then
960 bounced the address, did not kick in. This was a "< instead of <=" error;
961 in most cases the first failure would have been in the next clock tick
962 after the received time, and all would be well.
964 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
965 being recognized when the domain list was tested by the match_domain
966 condition in an expansion string.
968 16. Added the ${str2b64: operator.
970 17. Exim was always calling setrlimit() to set a large limit for the number of
971 processes, without checking whether the existing limit was already
972 adequate. (It did check for the limit on file descriptors.) Furthermore,
973 errors from getrlimit() and setrlimit() were being ignored. Now they are
974 logged to the main and panic logs, but Exim does carry on, to try to do its
975 job under whatever limits there are.
977 18. Imported PCRE 5.0.
979 19. Trivial typo in log message " temporarily refused connection" (the leading
982 20. If the log selector return_path_on_delivery was set and an address was
983 redirected to /dev/null, the delivery process crashed because it assumed
984 that a return path would always be set for a "successful" delivery. In this
985 case, the whole delivery is bypassed as an optimization, and therefore no
988 21. Internal re-arrangement: the function for sending a challenge and reading
989 a response while authentication was assuming a zero-terminated challenge
990 string. It's now changed to take a pointer and a length, to allow for
991 binary data in such strings.
993 22. Added the cyrus_sasl authenticator (code supplied by MBM).
995 23. Exim was not respecting finduser_retries when seeking the login of the
996 uid under which it was called; it was always trying 10 times. (The default
997 setting of finduser_retries is zero.) Also, it was sleeping after the final
998 failure, which is pointless.
1000 24. Implemented tls_on_connect_ports.
1002 25. Implemented acl_smtp_predata.
1004 26. If the domain in control=submission is set empty, Exim assumes that the
1005 authenticated id is a complete email address when it generates From: or
1006 Sender: header lines.
1008 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1009 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1010 chown and chgrp in /bin and hostname in /usr/bin.
1012 28. Exim was keeping the "process log" file open after each use, just as it
1013 does for the main log. This opens the possibility of it remaining open for
1014 long periods when the USR1 signal hits a daemon. Occasional processlog
1015 errors were reported, that could have been caused by this. Anyway, it seems
1016 much more sensible not to leave this file open at all, so that is what now
1019 29. The long-running daemon process does not normally write to the log once it
1020 has entered its main loop, and it closes the log before doing so. This is
1021 so that log files can straightforwardly be renamed and moved. However,
1022 there are a couple of unusual error situations where the daemon does write
1023 log entries, and I had neglected to close the log afterwards.
1025 30. The text of an SMTP error response that was received during a remote
1026 delivery was being truncated at 512 bytes. This is too short for some of
1027 the long messages that one sometimes sees. I've increased the limit to
1030 31. It is now possible to make retry rules that apply only when a message has a
1031 specific sender, in particular, an empty sender.
1033 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1034 it possible to be selective about when SMTP synchronization is enforced.
1036 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1038 32. Implemented hosts_connection_nolog.
1040 33. Added an ACL for QUIT.
1042 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1045 35. Added mailbox_size and mailbox_filecount to appendfile.
1047 36. Added control = no_multiline_responses to ACLs.
1049 37. There was a bug in the logic of the code that waits for the clock to tick
1050 in the case where the clock went backwards by a substantial amount such
1051 that the microsecond fraction of "now" was more than the microsecond
1052 fraction of "then" (but the whole seconds number was less).
1054 38. Added support for the libradius Radius client library this is found on
1055 FreeBSD (previously only the radiusclient library was supported).
1061 1. When certain lookups returned multiple values in the form name=value, the
1062 quoting of the values was not always being done properly. Specifically:
1063 (a) If the value started with a double quote, but contained no whitespace,
1065 (b) If the value contained whitespace other than a space character (i.e.
1066 tabs or newlines or carriage returns) it was not quoted.
1067 This fix has been applied to the mysql and pgsql lookups by writing a
1068 separate quoting function and calling it from the lookup code. The fix
1069 should probably also be applied to nisplus, ibase and oracle lookups, but
1070 since I cannot test any of those, I have not disturbed their existing code.
1072 2. A hit in the callout cache for a specific address caused a log line with no
1073 reason for rejecting RCPT. Now it says "Previous (cached) callout
1074 verification failure".
1076 3. There was an off-by-one bug in the queryprogram router. An over-long
1077 return line was truncated at 256 instead of 255 characters, thereby
1078 overflowing its buffer with the terminating zero. As well as fixing this, I
1079 have increased the buffer size to 1024 (and made a note to document this).
1081 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1082 when Exim is waiting for an SMTP response from a remote server, Exim
1083 restarts its select() call on the socket, thereby resetting its timeout.
1084 This is not a problem when such interrupts are rare. Somebody set up a cron
1085 job to run exiwhat every 2 minutes, which is less than the normal select()
1086 timeout (5 or 10 minutes). This meant that the select() timeout never
1087 kicked in because it was always reset. I have fixed this by comparing the
1088 time when an interrupt arrives with the time at the start of the first call
1089 to select(). If more time than the timeout has elapsed, the interrupt is
1090 treated as a timeout.
1092 5. Some internal re-factoring in preparation for the addition of Sieve
1093 extensions (by MH). In particular, the "personal" test is moved to a
1094 separate function, and given an option for scanning Cc: and Bcc: (which is
1095 not set for Exim filters).
1097 6. When Exim created an email address using the login of the caller as the
1098 local part (e.g. when creating a From: or Sender: header line), it was not
1099 quoting the local part when it contained special characters such as @.
1101 7. Installed new OpenBSD configuration files.
1103 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1104 try to make them clearer.
1106 9. Callout options, other than the timeout value, were being ignored when
1107 verifying sender addresses in header lines. For example, when using
1109 verify = header_sender/callout=no_cache
1111 the cache was (incorrectly) being used.
1113 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1114 only the Cygwin environment.
1116 11. When return_path_on_delivery was set as a log selector, if different remote
1117 addresses in the same message used different return paths and parallel
1118 remote delivery occurred, the wrong values would sometimes be logged.
1119 (Whenever a remote delivery process finished, the return path value from
1120 the most recently started remote delivery process was logged.)
1122 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1123 lines when AUTH and/or TLS are in use. This is the "received protocol"
1124 field. Exim used to use "asmtp" for authenticated SMTP, without any
1125 indication (in the protocol name) for TLS use. Now it follows the RFC and
1126 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1127 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1128 names appear in log lines as well as in Received: header lines.
1130 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1131 extensions, and comparison tests, and to fix some bugs.
1133 14. Changes to the "personal" filter test:
1135 (1) The test was buggy in that it was just doing the equivalent of
1136 "contains" tests on header lines. For example, if a user's address was
1137 anne@some.where, the "personal" test would incorrectly be true for
1139 To: susanne@some.where
1141 This test is now done by extracting each address from the header in turn,
1142 and checking the entire address. Other tests that are part of "personal"
1143 are now done using regular expressions (for example, to check local parts
1144 of addresses in From: header lines).
1146 (2) The list of non-personal local parts in From: addresses has been
1147 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1148 taken from the Sieve specification recommendations.
1150 (3) If the message contains any header line starting with "List-" it is
1151 treated as non-personal.
1153 (4) The test for "circular" in the Subject: header line has been removed
1154 because it now seems ill-conceived.
1156 15. Minor typos in src/EDITME comments corrected.
1158 16. Installed latest exipick from John Jetmore.
1160 17. If headers_add on a router specified a text string that was too long for
1161 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1162 of string_sprintf() is now avoided.
1164 18. $message_body_size was not set (it was always zero) when running the DATA
1165 ACL and the local_scan() function.
1167 19. For the "mail" command in an Exim filter, no default was being set for
1168 the once_repeat time, causing a random time value to be used if "once" was
1169 specified. (If the value happened to be <= 0, no repeat happened.) The
1170 default is now 0s, meaning "never repeat". The "vacation" command was OK
1171 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1172 (I found it when inspecting the code).
1174 20. There is now an overall timeout for performing a callout verification. It
1175 defaults to 4 times the callout timeout, which applies to individual SMTP
1176 commands during the callout. The overall timeout applies when there is more
1177 than one host that can be tried. The timeout is checked before trying the
1178 next host. This prevents very long delays if there are a large number of
1179 hosts and all are timing out (e.g. when the network connections are timing
1180 out). The value of the overall timeout can be changed by specifying an
1181 additional sub-option for "callout", called "maxwait". For example:
1183 verify = sender/callout=5s,maxwait=20s
1185 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1186 to the end before writing, but this should make it even safer).
1188 22. Exim was forgetting that it had advertised PIPELINING for the second and
1189 subsequent messages on an SMTP connection. It was also not resetting its
1190 memory on STARTTLS and an internal HELO.
1192 23. When Exim logs an SMTP synchronization error within a session, it now
1193 records whether PIPELINING has been advertised or not.
1195 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1196 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1197 rather than long int.
1199 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1201 26. Added the never_mail option to autoreply.
1207 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1208 crash if the getsockname() call failed; this can happen if a connection is
1209 closed very soon after it is established. The problem was simply in the
1210 order in which certain operations were done, causing Exim to try to write
1211 to the SMTP stream before it had set up the file descriptor. The bug has
1212 been fixed by making things happen in the correct order.
1218 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1219 before the connection was closed, thus losing the rejection response.
1221 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1222 some early Solaris releases, but causes trouble in current releases where
1223 socklen_t is defined.
1225 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1228 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1231 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1232 that releasing the top of it at the end releases what was used for sub-
1233 expansions (unless the block got too big). However, discard this block if
1234 the first thing is a variable or header, so that we can use its block when
1235 it is dynamic (useful for very large $message_headers, for example).
1237 6. Lookups now cache *every* query, not just the most recent. A new, separate
1238 store pool is used for this. It can be recovered when all lookup caches are
1239 flushed. Lookups now release memory at the end of their result strings.
1240 This has involved some general refactoring of the lookup sources.
1242 7. Some code has been added to the store_xxx() functions to reduce the amount
1243 of flapping under certain conditions.
1245 8. log_incoming_interface used to affect only the <= reception log lines. Now
1246 it causes the local interface and port to be added to several more SMTP log
1247 lines, for example "SMTP connection from", and rejection lines.
1249 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1251 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1253 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1254 could be overwritten at the end of the current message (or the start of a
1255 new message if it was set in a HELO ACL). The value is now preserved for
1256 the duration of the SMTP connection.
1258 12. If a transport had a headers_rewrite setting, and a matching header line
1259 contained an unqualified address, that address was qualified, even if it
1260 did not match any rewriting rules. The underlying bug was that the values
1261 of the flags that permit the existence of unqualified sender and recipient
1262 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1263 non-local messages, and by -bnq for local messages) were not being
1264 preserved with the message after it was received.
1266 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1267 "next input=" as part of the text comprising the host identity instead of
1268 the correct text. The code was using the same buffer for two different
1269 strings. However, depending on which order the printing function evaluated
1270 its arguments, the bug did not always show up. Under Linux, for example, my
1271 test suite worked just fine.
1273 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1274 doesn't work with some older versions of Perl. It has been changed to "my",
1275 which in any case is probably the better facility to use.
1277 15. A really picky compiler found some instances of statements for creating
1278 error messages that either had too many or two few arguments for the format
1281 16. The size of the buffer for calls to the DNS resolver has been increased
1282 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1283 for addresses that have a lot of PTR records. This alleviates a problem; it
1284 does not fully solve it.
1286 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1287 buffer now truncates the list and logs the incident, which is the same
1288 action as happens when Exim is looking up a host name and its aliases.
1289 Previously in this situation something unpredictable would happen;
1290 sometimes it was "internal error: store_reset failed".
1292 18. If a server dropped the connection unexpectedly when an Exim client was
1293 using GnuTLS and trying to read a response, the client delivery process
1294 crashed while trying to generate an error log message.
1296 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1297 string, for example:
1299 warn message = H1: something\nH2: something
1301 the text was added as a single header line from Exim's point of view
1302 though it ended up OK in the delivered message. However, searching for the
1303 second and subsequent header lines using $h_h2: did not work. This has been
1304 fixed. Similarly, if a system filter added multiple headers in this way,
1305 the routers could not see them.
1307 20. Expanded the error message when iplsearch is called with an invalid key to
1308 suggest using net-iplsearch in a host list.
1310 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1311 ACLs are no longer actually imposed (and a message to that effect is
1314 22. If a "gecos" field in a passwd entry contained escaped characters, in
1315 particular, if it contained a \" sequence, Exim got it wrong when building
1316 a From: or a Sender: header from that name. A second bug also caused
1317 incorrect handling when an unquoted " was present following a character
1318 that needed quoting.
1320 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1321 was not being matched caselessly.
1323 24. Arranged for all hyphens in the exim.8 source to be escaped with
1326 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1327 properly. Recipient callout cache records were still being keyed to include
1328 the sender, even when use_sender was set false. This led to far more
1329 callouts that were necessary. The sender is no longer included in the key
1330 when use_sender is false.
1332 26. Added "control = submission" modifier to ACLs.
1334 27. Added the ${base62d: operator to decode base 62 numbers.
1336 28. dnsdb lookups can now access SRV records.
1338 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1339 the configuration file.
1341 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1342 (-v) mode. This makes the output for a verbose queue run more intelligible.
1344 31. Added a use_postmaster feature to recipient callouts.
1346 32. Added the $body_zerocount variable, containing the number of binary zero
1347 bytes in the message body.
1349 33. The time of last modification of the "new" subdirectory is now used as the
1350 "mailbox time last read" when there is a quota error for a maildir
1353 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1355 35. Added +ignore_unknown as a special item in host lists.
1357 36. Code for decoding IPv6 addresses in host lists is now included, even if
1358 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1359 address was recognized as an IP address, but was then not correctly decoded
1360 into binary, causing unexpected and incorrect effects when compared with
1367 1. Very minor rewording of debugging text in manualroute to say "list of
1368 hosts" instead of "hostlist".
1370 2. If verify=header_syntax was set, and a header line with an unqualified
1371 address (no domain) and a large number of spaces between the end of the
1372 name and the colon was received, the reception process suffered a buffer
1373 overflow, and (when I tested it) crashed. This was caused by some obsolete
1374 code that should have been removed. The fix is to remove it!
1376 3. When running in the test harness, delay a bit after writing a bounce
1377 message to get a bit more predictability in the log output.
1379 4. Added a call to search_tidyup() just before forking a reception process. In
1380 theory, someone could use a lookup in the expansion of smtp_accept_max_
1381 per_host which, without the tidyup, could leave open a database connection.
1383 5. Added the variables $recipient_data and $sender_data which get set from a
1384 lookup success in an ACL "recipients" or "senders" condition, or a router
1385 "senders" option, similar to $domain_data and $local_part_data.
1387 6. Moved the writing of debug_print from before to after the "senders" test
1390 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1391 problems for message scanning, either using a data ACL, or using
1392 local_scan() because the Received: header was not generated till after they
1393 were called (in order to set the time as the time of reception completion).
1394 I have revised the way this works. The header is now generated after the
1395 body is received, but before the ACL or local_scan() are called. After they
1396 are run, the timestamp in the header is updated.
1402 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1403 before starting a queue runner without re-exec. This happened only when
1404 deliver_drop_privilege was set or when the Exim user was set to root. The
1405 effect of the bug was that timeouts during subsequent deliveries caused
1406 crashes instead of being properly handled. The handler is now left at its
1407 default (and expected) setting.
1409 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1410 message, again when deliver_drop_privilege is set or Exim is run as root.
1411 The bug described in (1) was not present in this case, but the tidying up
1412 of the other signals was missing. I have made the two cases consistent.
1414 3. The ignore_target_hosts setting on a manualroute router was being ignored
1415 for hosts that were looked up using the /MX notation.
1417 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1420 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1421 operated on the sender address. After changing the $sender_address to <>
1422 for the sender address verify, Exim was re-instated it as the original
1423 (before rewriting) address, but remembering that it had rewritten it, so it
1424 wasn't rewriting it again. This bug also had the effect of breaking the
1425 sender address verification caching when the sender address was rewritten.
1427 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1428 This has been changed so that if the ip literal address matches
1429 ignore_target_hosts, the router declines.
1431 7. Added expansion conditions match_domain, match_address, and match_local_
1432 part (NOT match_host).
1434 8. The placeholder for the Received: header didn't have a length field set.
1436 9. Added code to Exim itself and to exim_lock to test for a specific race
1437 condition that could lead to file corruption when using MBX delivery. The
1438 issue is with the lockfile that is created in /tmp. If this file is removed
1439 after a process has opened it but before that process has acquired a lock,
1440 there is the potential for a second process to recreate the file and also
1441 acquire a lock. This could lead to two Exim processes writing to the file
1442 at the same time. The added code performs the same test as UW imapd; it
1443 checks after acquiring the lock that its file descriptor still refers to
1444 the same named file.
1446 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1447 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1450 11. Added the smtp_active_hostname option. If used, this will typically be made
1451 to depend on the incoming interface address. Because $interface_address is
1452 not set up until the daemon has forked a reception process, error responses
1453 that can happen earlier (such as "too many connections") no longer contain
1456 12. If an expansion in a condition on a "warn" statement fails because a lookup
1457 defers, the "warn" statement is abandoned, and the next ACL statement is
1458 processed. Previously this caused the whole ACL to be aborted.
1460 13. Added the iplsearch lookup type.
1462 14. Added ident_timeout as a log selector.
1464 15. Added tls_certificate_verified as a log selector.
1466 16. Added a global option tls_require_ciphers (compare the smtp transport
1467 option of the same name). This controls incoming TLS connections.
1469 17. I finally figured out how to make tls_require_ciphers do a similar thing
1470 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1471 before starting the TLS session.
1473 18. Tabs are now shown as \t in -bP output.
1475 19. If the log selector return_path_on_delivery was set, Exim crashed when
1476 bouncing a message because it had too many Received: header lines.
1478 20. If two routers both had headers_remove settings, and the first one included
1479 a superfluous trailing colon, the final name in the first list and the
1480 first name in the second list were incorrectly joined into one item (with a
1481 colon in the middle).
1487 1. Added -C and -D options to the exinext utility, mainly to make it easier
1488 to include in the automated testing, but these could be helpful when
1489 multiple configurations are in use.
1491 2. The exinext utility was not formatting the output nicely when there was
1492 an alternate port involved in the retry record key, nor when there was a
1493 message id as well (for retries that were specific to a specific message
1494 and a specific host). It was also confused by IPv6 addresses, because of
1495 the additional colons they contain. I have fixed the IPv4 problem, and
1496 patched it up to do a reasonable job for IPv6.
1498 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1499 delivery, the log line now contains "pipelined" if PIPELINING was used.
1501 4. An SMTP transport process used to panic and die if the bind() call to set
1502 an explicit outgoing interface failed. This has been changed; it is now
1503 treated in the same way as a connect() failure.
1505 5. A reference to $sender_host_name in the part of a conditional expansion
1506 that was being skipped was still causing a DNS lookup. This no longer
1509 6. The def: expansion condition was not recognizing references to header lines
1510 that used bh_ and bheader_.
1512 7. Added the _cache feature to named lists.
1514 8. The code for checking quota_filecount in the appendfile transport was
1515 allowing one more file than it should have been.
1517 9. For compatibility with Sendmail, the command line option
1525 and sets the incoming protocol and host name (for trusted callers). The
1526 host name and its colon can be omitted when only the protocol is to be set.
1527 Note the Exim already has two private options, -pd and -ps, that refer to
1528 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1529 "s", but I don't think that's a major issue.
1531 10. A number of refactoring changes to the code, none of which should affect
1534 (a) The number of logging options was getting close to filling up the
1535 32-bit word that was used as a bit map. I have split them into two classes:
1536 those that are passed in the argument to log_write(), and those that are
1537 only ever tested independently outside of that function. These are now in
1538 separate 32-bit words, so there is plenty of room for expansion again.
1539 There is no change in the user interface or the logging behaviour.
1541 (b) When building, for example, log lines, the code previously used a
1542 macro that called string_cat() twice, in order to add two strings. This is
1543 not really sufficiently general. Furthermore, there was one instance where
1544 it was actually wrong because one of the argument was used twice, and in
1545 one call a function was used. (As it happened, calling the function twice
1546 did not affect the overall behaviour.) The macro has been replaced by a
1547 function that can join an arbitrary number of extra strings onto a growing
1550 (c) The code for expansion conditions now uses a table and a binary chop
1551 instead of a serial search (which was left over from when there were very
1552 few conditions). Also, it now recognizes conditions like "pam" even when
1553 the relevant support is not compiled in: a suitably worded error message is
1554 given if an attempt is made to use such a condition.
1556 11. Added ${time_interval:xxxxx}.
1558 12. A bug was causing one of the ddress fields not to be passed back correctly
1559 from remote delivery subprocesses. The field in question was not being
1560 subsequently used, so this caused to problems in practice.
1562 13. Added new log selectors queue_time and deliver_time.
1564 14. Might have fixed a bug in maildirsizefile handling that threw up
1565 "unexpected character" debug warnings, and recalculated the data
1566 unnecessarily. In any case, I expanded the warning message to give more
1569 15. Added the message "Restricted characters in address" to the statements in
1570 the default ACL that block characters like @ and % in local parts.
1572 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1573 Three changes have been made:
1575 (a) There was a serious bug; a negative response to MAIL caused the whole
1576 recipient domain to be cached as invalid, thereby blocking all messages
1577 to all local parts at the same domain, from all senders. This bug has
1578 been fixed. The domain is no longer cached after a negative response to
1579 MAIL if the sender used is not empty.
1581 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1584 (c) A new callout option, "use_sender" has been added for people who want
1585 the modified behaviour.
1591 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1594 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1595 indeed breaks things for older releases.
1597 3. Added additional logging to the case where there is a problem reading data
1598 from a filter that is running in a subprocess using a pipe, in order to
1599 try to track down a specific problem.
1601 4. Testing facility fudge: when running in the test harness and attempting
1602 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1603 getting "No route to host". Convert this to a timeout.
1605 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1608 6. Some OS don't have socklen_t but use size_t instead. This affects the
1609 fifth argument of getsockopt() amongst other things. This is now
1610 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1611 can be set for individual OS. I have set it for SunOS5, OSF1, and
1612 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1613 some earlier ones do not.
1615 7. Change 4.30/15 was not doing the test caselessly.
1617 8. The standard form for an IPv6 address literal was being rejected by address
1618 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1619 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1620 this, as well as the form without the "IPv6" on the front (but only when
1621 address literals are enabled, of course).
1623 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1625 10. Exim crashed if a message with an empty sender address specified by -f
1626 encountered a router with an errors_to setting. This could be provoked only
1627 by a command such as
1631 where an empty string was supplied; "<>" did not hit this bug.
1633 11. Installed PCRE release 4.5.
1635 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1636 remained set. It is now erased.
1638 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1639 times from message ids (which are base 36 rather than the normal 62).
1641 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1642 were being counted as actual protocol errors, and logged if the log
1643 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1644 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1645 and DATA following a set of rejected RCPTs do not count as protocol errors.
1646 In other words, Exim assumes they were pipelined, though this may not
1647 actually be the case. Of course, in all cases the client gets an
1648 appropriate error code.
1650 15. If a lookup fails in an ACL condition, a message about the failure may
1651 be available; it is used if testing the ACL cannot continue, because most
1652 such messages specify what the cause of the deferral is. However, some
1653 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1654 that caused an old message to be retained and used if a later statement
1655 caused a defer, replacing the real cause of the deferral.
1657 16. If an IP address had so many PTR records that the DNS lookup buffer
1658 was not large enough to hold them, Exim could crash while trying to process
1659 the truncated data. It now detects and logs this case.
1661 17. Further to 4.21/58, another change has been made: if (and only if) the
1662 first line of a message (the first header line) ends with CRLF, a bare LF
1663 in a subsequent header line has a space inserted after it, so as not to
1664 terminate the header.
1666 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1667 unnecessarily, used atoi() instead of strtol(), and didn't check the
1668 termination when getting file sizes from file names by regex.
1670 19. Completely re-implemented the support for maildirsize files, in the light
1671 of a number of problems with the previous contributed implementation
1672 (4.30/29). In particular:
1674 . If the quota is zero, the maildirsize file is maintained, but no quota is
1677 . If the maildir directory does not exist, it is created before any attempt
1678 to write a maildirsize file.
1680 . The quota value in the file is just a cache; if the quota is changed in
1681 the transport, the new value overrides.
1683 . A regular expression is available for excluding directories from the
1686 20. The autoreply transport checks the characters in options that define the
1687 message's headers; it allows continued headers, but it was checking with
1688 isspace() after an embedded newline instead of explicitly looking for a
1691 21. If all the "regular" hosts to which an address was routed had passed their
1692 expiry times, and had not reached their retry times, the address was
1693 bounced, even if fallback hosts were defined. Now Exim should go on to try
1696 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1697 equivalent code in the SMTP transport. Some hosts send humungous responses
1698 to HELO/EHLO, more than 1024 it seems.
1700 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1701 gives compiler warnings in some environments. I've now done it "properly",
1704 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1705 (because of problems with the built-in one) was declared to return uschar *
1706 instead of char *, causing compiler failure.
1708 25. Fixed a file descriptor leak when processing alias/forward files.
1710 26. Fixed a minor format string issue in dbfn.c.
1712 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1714 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1715 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1718 29. When a sender address is verified, it is cached, to save repeating the test
1719 when there is more than one recipient in a message. However, when the
1720 verification involves a callout, it is possible for different callout
1721 options to be set for different recipients. It is too complicated to keep
1722 track of this in the cache, so now Exim always runs a verification when a
1723 callout is required, relying on the callout cache for the optimization.
1724 The overhead is duplication of the address routing, but this should not be
1727 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1728 to be verified with callout=postmaster, and the main callout worked but the
1729 postmaster check failed, the verification correctly failed. However, if a
1730 subsequent RCPT command asked for sender verification *without* the
1731 postmaster check, incorrect caching caused this verification also to fail,
1734 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1735 it was not caching the DNS options (qualify_single, search_parents) that
1736 were used when the lookup failed. A subsequent lookup with different
1737 options therefore always gave the same answer, though there were cases
1738 where it should not have. (Example: a "domains = !$mx_any" option on a
1739 dnslookup router: the "domains" option is always processed without any
1740 widening, but the router might have qualify_single set.) Now Exim uses the
1741 cached value only when the same options are set.
1743 32. Added John Jetmore's "exipick" utility to the distribution.
1745 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1746 than a timeout (e.g. a certificate is required, and is not provided), an
1747 Exim server now closes the connection immediately. Previously it waited for
1748 the client to close - but if the client is SSL, it seems that they each
1749 wait for each other, leading to a delay before one of them times out.
1751 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1752 maintained 0.8.x compatibility because I don't think many are using it, and
1753 it is clearly obsolete.
1755 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1758 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1759 client certificate was expired. A simple patch fixes this, though I don't
1760 understand the full logic of why the verify callback is called multiple
1763 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1764 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1765 which causes problems with some clients (such as the Certicom SSL Plus
1766 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1767 disables the coutermeasure allowing Eudora to connect."
1769 38. Exim was not checking that a write() to a log file succeeded. This could
1770 lead to Bad Things if a log got too big, in particular if it hit a file
1771 size limit. Exim now panics and dies if it cannot write to a log file, just
1772 as it does if it cannot open a log file.
1774 39. Modified OS/Makefile-Linux so that it now contains
1776 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1778 The two -D definitions ensure that Exim is compiled with large file
1779 support, which makes it possible to handle log files that are bigger than
1782 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1783 instance) a domain was checked against a named list that involved a lookup,
1784 causing $domain_data to be set, then another domain was checked against the
1785 same list, then the first domain was re-checked, the value of $domain_data
1786 after the final check could be wrong. In particular, if the second check
1787 failed, it could be set empty. This bug probably also applied to
1790 41. The strip_trailing_dot option was not being applied to the address given
1791 with the -f command-line option.
1793 42. The code for reading a message's header from the spool was incrementing
1794 $received_count, but never initializing it. This meant that the value was
1795 incorrect (doubled) while delivering a message in the same process in which
1796 it was received. In the most common configuration of Exim, this never
1797 happens - a fresh exec is done - but it can happen when
1798 deliver_drop_privilege is set.
1800 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1801 it now includes up to 150 characters of the unexpected data in the log
1804 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1805 and building data strings. The size of both of these buffers was 10 000
1806 bytes - far larger than anybody would *ever* want, thought I. Needless to
1807 say, somebody hit the limit. I have increased the maximum line length to
1808 20 000 and the maximum data length of concatenated lines to 100 000. I have
1809 also fixed two bugs, because there was no checking on these buffers. Tsk,
1810 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1811 buffer is too small.
1813 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1814 lsearch lookups. Now it does.
1816 46. When parsing a route_list item in a manualroute router, a fixed-length
1817 buffer was used for the list of hosts. I made this 1024 bytes long,
1818 thinking that nobody would ever have a list of hosts that long. Wrong.
1819 Somebody had a whole pile of complicated expansion conditions, and the
1820 string was silently truncated, leading to an expansion error. It turns out
1821 that it is easier to change to an unlimited length (owing to other changes
1822 that have happened since this code was originally written) than to build
1823 structure for giving a limitation error. The length of the item that
1824 expands into the list of hosts is now unlimited.
1826 47. The lsearch lookup could not handle data where the length of text line was
1827 more than 4095 characters. Such lines were truncated, leading to shortened
1828 data being returned. It should now handle lines of any length.
1830 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1831 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1834 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1835 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1836 no longer happens in comment lines. A list of replacements is now placed
1837 at the head of all of the source files, except those whose only change is
1838 to replace PERL_COMMAND in the very first #! line.
1840 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1841 messages on the queue, with a bottom-up merge sort, using code contributed
1842 by Michael Haardt. This should make operations like -bp somewhat faster on
1843 large queues. It won't affect queue runners, except when queue_run_in_order
1846 51. Installed eximstats 1.31 in the distribution.
1848 52. Added support for SRV lookups to the dnslookup router.
1850 53. If an ACL referred to $message_body or $message_body_end, the value was not
1851 reset for any messages that followed in the same SMTP session.
1853 54. The store-handling optimization for building very long strings was not
1854 differentiating between the different store pools. I don't think this
1855 actually made any difference in practice, but I've tidied it.
1857 55. While running the routers to verify a sender address, $sender_address
1858 was still set to the sender address. This is wrong, because when routing to
1859 send a bounce to the sender, it would be empty. Therefore, I have changed
1860 it so that, while verifying a sender address, $sender_address is set to <>.
1861 (There is no change to what happens when verifying a recipient address.)
1863 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1864 target A or AAAA records (if not already returned) without resetting the
1865 qualify_single or search_parents options of the DNS resolver. These are
1866 inappropriate in this case because the targets of MX and SRV records must
1867 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1868 target that, when qualified, matched something in the local domain. These
1869 two options are now turned off when doing these lookups.
1871 57. It seems that at least some releases of Reiserfs (which does not have the
1872 concept of a fixed number of inodes) returns zero and not -1 for the
1873 number of available inodes. This interacted badly with check_spool_inodes,
1874 which assumed that -1 was the "no such thing" setting. What I have done is
1875 to check that the total number of inodes is greater than zero before doing
1876 the test of how many are available.
1878 58. When a "warn" ACL statement has a log_message modifier, the message is
1879 remembered, and not repeated. This is to avoid a lot of repetition when a
1880 message has many recipients that cause the same warning to be written.
1881 Howewer, Exim was preserving the list of already written lines for an
1882 entire SMTP session, which doesn't seem right. The memory is now reset if a
1883 new message is started.
1885 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1886 debugging output unless log_rewrite was also set.
1888 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1889 of (int)(handle) when we know that handle contains (void *)(-1).
1891 61. The Exim daemon panic-logs an error return when it closes the incoming
1892 connection. However "connection reset by peer" seems to be common, and
1893 isn't really an error worthy of noting specially, so that particular error
1896 62. When Exim is trying to find all the local interfaces, it used to panic and
1897 die if the ioctl to get the interface flags failed. However, it seems that
1898 on at least one OS (Solaris 9) it is possible to have an interface that is
1899 included in the list of interfaces, but for which you get a failure error
1900 for this call. This happens when the interface is not "plumbed" into a
1901 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1902 failure of the "get flags" call assumes that the interface is down.
1904 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1905 makes life easier for people who are doing arithmetic on fields extracted
1906 from dates, where you often get leading zeros that should not be
1907 interpreted as octal.
1909 64. Added qualify_domain to the redirect router, to override the global
1912 65. If a pathologically long header line contained very many addresses (the
1913 report of this problem mentioned 10 000) and each of them was rewritten,
1914 Exim could use up a very large amount of memory. (It kept on making new
1915 copies of the header line as it rewrote, and never released the old ones.)
1916 At the expense of a bit more processing, the header rewriting function has
1917 been changed so that it no longer eats memory in this way.
1919 66. The generation of the Received: header has been moved from the time that a
1920 message starts to be received, to the time that it finishes. The timestamp
1921 in the Received: header should now be very close to that of the <= log
1922 line. There are two side-effects of this change:
1924 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1925 logged header lines no longer include the local Received: line, because
1926 it has not yet been created. The same applies to a copy of the message
1927 that is returned to a non-SMTP sender when a message is rejected.
1929 (b) When a filter file is tested using -bf, no additional Received: header
1930 is added to the test message. After some thought, I decided that this
1933 This change does not affect the value of $received_for. It is still set
1934 after address rewriting, but before local_scan() is called.
1936 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1938 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1939 gave an unhelpful panic error message, and a defer error. I have managed to
1940 change this behaviour so that it now rejects any supplied certificate,
1941 which seems right, as the list of acceptable certificates is empty.
1943 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1944 gave an unhelpful defer error. I have not managed to make this reject any
1945 supplied certificates, but the error message it gives is "no certificate
1946 supplied", which is not helpful.
1948 70. exigrep's output now also includes lines that are not associated with any
1949 message, but which match the given pattern. Implemented by a patch from
1950 Martin Sluka, which also tidied up the Perl a bit.
1952 71. Recipient callout verification, like sender verification, was using <> in
1953 the MAIL FROM command. This isn't really the right thing, since the actual
1954 sender may affect whether the remote host accepts the recipient or not. I
1955 have changed it to use the actual sender in the callout; this means that
1956 the cache record is now keyed on a recipient/sender pair, not just the
1957 recipient address. There doesn't seem to be a real danger of callout loops,
1958 since a callout by the remote host to check the sender would use <>.
1959 [SEE ABOVE: changed after hitting problems.]
1961 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1962 temporary errors. However, in the case of such a code being given after
1963 the end of a data transmission (i.e. after ".") Exim was failing to write
1964 a retry record for the message. (Yes, there was some broken host that was
1965 actually sending 8xx at this point.)
1967 73. An unknown lookup type in a host list could cause Exim to panic-die when
1968 the list was checked. (An example that provoked this was putting <; in the
1969 middle of a list instead of at the start.) If this happened during a DATA
1970 ACL check, a -D file could be left lying around. This kind of configuration
1971 error no longer causes Exim to die; instead it causes a defer errror. The
1972 incident is still logged to the main and panic logs.
1974 74. Buglet left over from Exim 3 conversion. The message "too many messages
1975 in one connection" was written to the rejectlog but not the mainlog, except
1976 when address rewriting (yes!) was being logged.
1978 75. Added write_rejectlog option.
1980 76. When a system filter was run not as root (that is, when system_filter_user
1981 was set), the values of the $n variables were not being returned to the
1982 main process; thus, they were not subsequently available in the $sn
1985 77. Added +return_path_on_delivery log selector.
1987 78. A connection timeout was being treated differently from recipients deferred
1988 when testing hosts_max_try with a message that was older than the host's
1989 retry timeout. (The host should not be counted, thus allowing all hosts to
1990 be tried at least once before bouncing.) This may have been the cause of an
1991 occasionally reported bug whereby a message would remain on the queue
1992 longer than the retry timeout, but would be bounced if a delivery was
1993 forced. I say "may" because I never totally pinned down the problem;
1994 setting up timeout/retry tests is difficult. See also the next item.
1996 79. The ultimate address timeout was not being applied to errors that involved
1997 a combination of host plus message (for example, a timeout on a MAIL
1998 command). When an address resolved to a number of possible hosts, and they
1999 were not all tried for each delivery (e.g. because of hosts_max_try), a
2000 message could remain on the queue longer than the retry timeout.
2002 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2005 81. Fixed an obscure SMTP outgoing bug which required at least the following
2006 conditions: (a) there was another message waiting for the same server;
2007 (b) the server returned 5xx to all RCPT commands in the first message so
2008 that the message was not completed; (c) the server dropped the connection
2009 or gave a negative response to the RSET that Exim sends to abort the
2010 transaction. The observed case was a dropped connection after DATA that had
2011 been sent in pipelining mode. That is, the server had advertised PIPELINING
2012 but was not implementing it correctly. The effect of the bug was incorrect
2013 behaviour, such as trying another host, and this could lead to a crash.
2019 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2020 and daemon.c were passed as pointers to ints; they should have been
2021 pointers to socklen_t variables (which are typically unsigned ints).
2023 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2026 3. Fixed a really odd bug that affected only the testing scheme; patching a
2027 certain fixed string in the binary changed the value of another string that
2028 happened to be identical to the end of the original first string.
2030 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2031 name", it returns that address as the IP address. On some operating
2032 systems (e.g. Solaris), it also passes back the IP address string as the
2033 "host name". However, on others (e.g. Linux), it passes back an empty
2034 string. Exim wasn't checking for this, and was changing the host name to an
2035 empty string, assuming it had been canonicized.
2037 5. Although rare, it is permitted to have more than one PTR record for a given
2038 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2039 all the names associated with an address, because they do in Solaris.
2040 However, it seems that they do not in Linux for data that comes from the
2041 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2042 I found this out when I moved to a new Linux workstation and tried to run
2043 the Exim test suite.
2045 To get round this problem I have changed the code so that it now does its
2046 own call to the DNS to look up PTR records when searching for a host name.
2047 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2048 addresses that are only in /etc/hosts are still found.
2050 This behaviour is, however, controlled by an option called host_lookup_
2051 order, which defaults to "bydns:byaddr". If people want to use the other
2052 order, or indeed, just use one or the other means of lookup, they can
2053 specify it in this variable.
2055 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2056 some operating systems, this comes back from gethostbyaddr() as an empty
2057 string, and this is what Exim used to test for. However, it seems that in
2058 other systems, "." is yielded. Exim now tests for this case too.
2060 7. The values of check_spool_space and check_log_space are now held internally
2061 as a number of kilobytes instead of an absolute number of bytes. If a
2062 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2063 kilobyte. This means that much larger values can be stored.
2065 8. Exim monitor: an attempt to get the action menu when not actually pointing
2066 at a message produces an empty menu entitled "No message selected". This
2067 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2068 no entries in it ("Shell widget menu has zero width and/or height"). So I
2069 have added a single, blank menu entry in this case.
2071 9. Added ${quote_local_part.
2073 10. MIME decoding is now applied to the contents of Subject: header lines when
2076 11. Now that a reference to $sender_host_address automatically causes a reverse
2077 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2078 host lookup before query-style lookups in lists that might use this
2079 variable. This has therefore been abolished, and the "net-" prefix is no
2080 longer necessary for query-style lookups.
2082 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2083 have been a typo for LFLAGS, so it has been changed.
2085 13. The install script calls Exim with "-C /dev/null" in order to find the
2086 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2087 to be output. Howeve, since Exim outputs its version number before the
2088 error, it didn't break the script. It just looked ugly. I fixed this by
2089 always allowing "-C /dev/null" if the caller is root.
2091 14. Ignore overlarge ACL variable number when reading spool file - insurance
2092 against a later release with more variables having written the file.
2094 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2095 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2096 this, as well as the form without the "IPv6" on the front.
2098 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2099 OS/Makefile-Darwin file.
2101 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2102 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2104 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2105 variable. However, some OS use ints for inodes. I've added cast to long int
2106 to get rid of the compiler warning.
2108 19. I had forgotten to lock out "/../" in configuration file names when
2109 ALT_CONFIG_PREFIX was set.
2111 20. Routers used for verification do not need to specify transports. However,
2112 if such a router generated a host list, and callout was configured, Exim
2113 crashed, because it could not find a port number from the (non-existent)
2114 transport. It now assumes port 25 in this circumstance.
2116 21. Added the -t option to exigrep.
2118 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2119 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2120 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2121 them is set, LOOKUP_LSEARCH is forced.
2123 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2125 24. Added sender and host information to the "rejected by local_scan()" log
2126 line; previously there was no indication of these.
2128 25. Added .include_if_exists.
2130 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2131 Linux. It turns out that not all file systems support this. Apparently some
2132 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2133 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2134 not supported on the file descriptor, is now ignored when Exim is trying to
2135 sync a directory. This applies only to Linux.
2137 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2139 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2140 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2143 29. Added support for maildirsize files from supplied patch (modified a bit).
2145 30. The use of :fail: followed by an empty string could lead Exim to respond to
2146 sender verification failures with (e.g.):
2148 550 Verification failed for <xxx>
2149 550 Sender verify failed
2151 where the first response line was missing the '-' that indicates it is not
2152 the final line of the response.
2154 31. The loop for finding the name of the user that called Exim had a hardwired
2155 limit of 10; it now uses the value of finduser_retries, which is used for
2156 all other user lookups.
2158 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2161 33. Exim was neglecting to zero errno before one call of strtol() when
2162 expanding a string and expecting an integer value. On some systems this
2163 resulted in spurious "integer overflow" errors. Also, it was casting the
2164 result into an int without checking.
2166 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2167 did not work. The code looks as if it has *never* worked, though it appears
2168 to have been documented since at least releast 1.62. I have made it work.
2170 35. The "timeout_DNS" error in retry rules, also documented since at least
2171 1.62, also never worked. As it isn't clear exactly what this means, and
2172 clearly it isn't a major issue, I have abolished the feature by treating it
2173 as "timeout", and writing a warning to the main and panic logs.
2175 36. The display of retry rules for -brt wasn't always showing the error code
2178 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2179 timeout_connect_A, timeout_connect_MX.
2181 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2182 to the empty sender.
2184 39. The daemon was not analysing the content of -oX till after it had closed
2185 stderr and disconnected from the controlling terminal. This meant that any
2186 syntax errors were only noted on the panic log, and the return code from
2187 the command was 0. By re-arranging the code a little, I've made the
2188 decoding happen first, so such errors now appear on stderr, and the return
2189 code is 1. However, the actual setting up of the sockets still happens in
2190 the disconnected process, so errors there are still only recorded on the
2193 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2194 connections (as happens on some IP stacks) was logged at start up time as
2195 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2196 it from "IPv6 and IPv4", which means that two separate sockets are being
2199 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2200 says whether AF_INET or AF_INET6 was passed as an argument.
2202 42. Exiwhat output was messed up when time zones were included in log
2205 43. Exiwhat now gives more information about the daemon's listening ports,
2206 and whether -tls-on-connect was used.
2208 44. The "port" option of the smtp transport is now expanded.
2210 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2211 silently ignored. Now an error message is written to the main and panic
2214 46. There's a new ACL modifier called "logwrite" which writes to a log file
2215 as soon as it is encountered.
2217 47. Added $local_user_uid and $local_user_gid at routing time.
2219 48. Exim crashed when trying to verify a sender address that was being
2222 49. Exim was recognizing only a space character after ".include". It now also
2223 recognizes a tab character.
2225 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2226 extracting the relevant information from the specification. The man page no
2227 longer contains scrambled data for the -d option, and I've added a section
2228 at the front about calling Exim under different names.
2230 51. Added "extra_headers" argument to the "mail" command in filter files.
2232 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2235 53. Installed eximstats 1.29.
2237 54. Added transport_filter_timeout as a generic transport option.
2239 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2240 Cc: header lines. This was required by RFC 822, but it not required by RFC
2243 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2244 incoming messages that did not have them. Now it does so only if the
2245 message originates locally, that is, if there is no associated remote host
2246 address. When Resent- header lines are present, this applies to the Resent-
2247 lines rather than the non-Resent- lines.
2249 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2250 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2252 58. Messages for configuration errors now include the name of the main
2253 configuration file - useful now that there may be more than one file in a
2254 list (.included file names were always shown).
2256 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2257 for those rare installations that do not start the daemon as root or run it
2258 setuid root. I've cut out the call to initgroups() if the daemon is not
2261 60. The Exim user and group can now be bound into the binary as text strings
2262 that are looked up at the start of Exim's processing.
2264 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2266 62. Added $mailstore_basename variable.
2268 63. Installed patch to sieve.c from Michael Haardt.
2270 64. When Exim failed to open the panic log after failing to open the main log,
2271 the original message it was trying to log was written to stderr and debug
2272 output, but if they were not available (the usual case in production), it
2273 was lost. Now it is written to syslog before the two lines that record the
2274 failures to open the logs.
2276 65. Users' Exim filters run in subprocesses under the user's uid. It is
2277 possible for a "deliver" command or an alias in a "personal" command to
2278 provoke an address rewrite. If logging of address rewriting is configured,
2279 this fails because the process is not running as root or exim. There may be
2280 a better way of dealing with this, but for the moment (because 4.30 needs
2281 to be released), I have disabled address rewrite logging when running a
2282 filter in a non-root, non-exim process.
2288 1. The buildconfig auxiliary program wasn't quoting the value set for
2289 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2290 not defined. This bug was masked in 4.22 by the effect that was fixed in
2293 2. Some messages that were rejected after a message id was allocated were
2294 shown as "incomplete" by exigrep. It no longer does this for messages that
2295 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2297 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2298 have allow_domain_literals set, the ID did not get logged in the <= line.
2299 Domain literals are now always recognized in Message-ID: header lines.
2301 4. The first argument for a ${extract expansion item is the key name or field
2302 number. Leading and trailing spaces in this item were not being ignored,
2303 causing some misleading effects.
2305 5. When deliver_drop_privilege was set, single queue runner processes started
2306 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2307 same command in the process it spins off) were not dropping privilege.
2309 6. When the daemon running as "exim" started a queue runner, it always
2310 re-executed Exim in the spun-off process. This is a waste of effort when
2311 deliver_drop_privilege is set. The new process now just calls the
2312 queue-runner function directly.
2318 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2321 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2322 ignored. Though the use of -oP was forcing the writing of a pid file, it
2323 was always written to the default place.
2325 3. If the message "no IP address found for host xxxx" is generated during
2326 incoming verification, it is now followed by identification of the incoming
2327 connection (so you can more easily find what provoked it).
2329 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2331 5. Added some features to "harden" Exim a bit more against certain attacks:
2333 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2334 be put in Local/Makefile. This is like the never_users runtime option,
2335 but it cannot be overridden. The default setting is "root".
2337 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2338 prefix string with which any file named in a -C command line option
2341 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2342 is retained for -C and -D only if the caller of Exim is root. Without
2343 it, the exim user may also use -C and -D and retain privilege.
2345 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2346 command line option is disabled.
2348 6. Macro names set by the -D option must start with an upper case letter, just
2349 like macro names defined in the configuration file.
2351 7. Added "dereference=" facility to LDAP.
2353 8. Two instances of the typo "uknown" in the source files are fixed.
2355 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2356 the Configure-Makefile script screwed up while processing it.
2358 10. Incorporated PCRE 4.4.
2360 11. The SMTP synchronization check was not operating right at the start of an
2361 SMTP session. For example, it could not catch a HELO sent before the client
2362 waited for the greeting. There is now a check for outstanding input at the
2363 point when the greeting is written. Because of the duplex, asynchronous
2364 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2365 way, but not yet received, when the check is performed.
2367 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2368 on TCP/IP sockets, because this apparently causes some broken clients to
2371 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2372 unchanged) from the Cygwin maintainer.
2374 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2375 was supported instead of testing for mbx. Effectively a typo.
2377 15. The spa authenticator server code was not checking that the input it
2378 received was valid base64.
2380 16. The debug output line for the "set" modifier in ACLs was not showing the
2381 name of the variable that was being set.
2383 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2385 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2386 lookup on its own. Something else was needed to trigger the lookup. For
2387 example, a match in host_lookup or the need for a host name in a host list.
2388 Now, if $sender_host_name is referenced and the host name has not yet been
2389 looked up, a lookup is performed. If the lookup fails, the variable remains
2390 empty, and $host_lookup_failed is set to "1".
2392 19. Added "eqi" as a case-independent comparison operator.
2394 20. The saslauthd authentication condition could segfault if neither service
2395 nor realm was specified.
2397 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2398 error message that was logged was misleading, and incoming SMTP
2399 connections were dropped. The message is now more accurate, and temporary
2400 errors are given to SMTP connections.
2402 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2403 (except RSET) with a 421 error, until QUIT is received. However, it was
2404 failing to send a response to QUIT.
2406 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2407 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2408 ACL is now run after the helo_try_verify_hosts code.
2410 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2411 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2412 case-independent, so other case variants are also recognized). Apparently
2413 some systems use these upper case variants.
2415 25. If more than two messages were waiting for the same host, and a transport
2416 filter was specified for the transport, Exim sent two messages over the
2417 same TCP/IP connection, and then failed with "socket operation on non-
2418 socket" when it tried to send the third.
2420 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2422 27. The extern definition of crypt16() in expand.c was not being excluded when
2423 the OS had its own crypt16() function.
2425 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2426 as a preferred synonym for return_size_limit, both as an option and as an
2429 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2431 30. Changed the default configuration ACL to relax the local part checking rule
2432 for addresses that are not in any local domains. For these addresses,
2433 slashes and pipe symbols are allowed within local parts, but the sequence
2434 /../ is explicitly forbidden.
2436 31. SPA server authentication was not clearing the challenge buffer before
2439 32. log_message in a "warn" ACL statement was writing to the reject log as
2440 well as to the main log, which contradicts the documentation and doesn't
2441 seem right (because no rejection is happening). So I have stopped it.
2443 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2444 However, I am unable to do any testing of this.
2446 34. Fixed an infelicity in the appendfile transport. When checking directories
2447 for a mailbox, to see if any needed to be created, it was accidentally
2448 using path names with one or more superfluous leading slashes; tracing
2449 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2451 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2452 contents are added to the log line that is written for every discarded
2453 recipient. (Previously a log_message setting was ignored.)
2455 36. The ${quote: operator now quotes the string if it is empty.
2457 37. The install script runs exim in order to find its version number. If for
2458 some reason other than non-existence or emptiness, which it checks, it
2459 could not run './exim', it was installing it with an empty version number,
2460 i.e. as "exim-". This error state is now caught, and the installation is
2463 38. An argument was missing from the function that creates an error message
2464 when Exim fails to connect to the socket for saslauthd authentication.
2465 This could cause Exim to crash, or give a corrupted message.
2467 39. Added isip, isip4, and isip6 to ${if conditions.
2469 40. The ACL variables $acl_xx are now saved with the message, and can be
2470 accessed later in routers, transports, and filters.
2472 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2473 strings in the file are not string-expanded.
2475 42. If a MAIL command specified a SIZE value that was too large to fit into an
2476 int variable, the check against message_size_limit failed. Such values are
2477 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2478 day this will have to be increased, but I don't think I want to be around
2479 when emails are that large.
2486 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2487 iconv() is not standard in FreeBSD.
2489 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2490 IPv6 enabled. The observed symptom was a segmentation fault on return from
2491 the function os_common_find_running_interfaces() in src/os.c.
2493 3. In the check_special_case() function in daemon.c I had used "errno" as an
2494 argument name, which causes warnings on some systems. This was basically a
2495 typo, since it was named "eno" in the comments!
2497 4. The code that waits for the clock to tick (at a resolution of some fraction
2498 of a second) so as to ensure message-id uniqueness was always waiting for
2499 at least one whole tick, when it could have waited for less. [This is
2500 almost certainly not relevant at current processor speeds, where it is
2501 unlikely to ever wait at all. But we try to future-proof.]
2503 5. The function that sleeps for a time interval that includes fractions of a
2504 second contained a race. It did not block SIGALRM between setting the
2505 timer, and suspending (a couple of lines later). If the interval was short
2506 and the sigsuspend() was delayed until after it had expired, the suspension
2507 never ended. On busy systems this could lead to processes getting stuck for
2510 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2511 process, before it forks any delivery processes. The open lookup caching
2512 mechanism meant that the open file or database connection was passed into
2513 the delivery process. The problem was that delivery processes always tidy
2514 up cached lookup data. This could cause a problem for the next delivery
2515 process started by the queue runner, because the external queue runner
2516 process does not know about the closure. So the next delivery process
2517 still has data in the lookup cache. In the case of a file lookup, there was
2518 no problem because closing a file descriptor in a subprocess doesn't affect
2519 the parent. However, if the lookup was caching a connection to a database,
2520 the connection was closed, and the second delivery process was likely to
2521 see errors such as "PGSQL: query failed: server closed the connection
2522 unexpectedly". The problem has been fixed by closing all cached lookups
2523 in a queue runner before running a delivery process.
2525 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2526 seem to have the "const" qualifier which it has on other OS. I've
2529 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2530 *of the same type* (client or server) with the same public name that an
2531 error should be diagnosed.
2533 9. When Exim looked up a host name for an IP address, but failed to find the
2534 original IP address when looking up the host name (a safety check), it
2535 output the message "<ip address> does not match any IP for NULL", which was
2536 confusing, to say the least. The bug was that the host name should have
2537 appeared instead of "NULL".
2539 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2540 user that is built into the binary, and the -C or -D options is used, root
2541 privilege is dropped before the configuration file is read. In addition,
2542 logging is switched to stderr instead of the normal log files. If the
2543 configuration then re-defines the Exim user, the unprivileged environment
2544 is probably not what is expected, so Exim logs a panic warning message (but
2547 However, if deliver_drop_privilege is set, the unprivileged state may well
2548 be exactly what is intended, so the warning has been cut out in that case,
2549 and Exim is allowed to try to write to its normal log files.
2555 1. smtp_return_error_details was not giving details for temporary sender
2556 or receiver verification errors.
2558 2. Diagnose a configuration error if two authenticators have the same public
2561 3. Exim used not to create the message log file for a message until the first
2562 delivery attempt. This could be confusing when incoming messages were held
2563 for policy or load reasons. The message log file is now created at the time
2564 the message is received, and an initial "Received" line is written to it.
2566 4. The automatically generated man page for command line options had a minor
2567 bug that caused no ill effects; however, a more serious problem was that
2568 the procedure for building the man page automatically didn't always
2569 operate. Consequently, release 4.20 contains an out-of-date version. This
2570 shouldn't happen again.
2572 5. When building Exim with embedded Perl support, the script that builds the
2573 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2574 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2576 6. The freeze_tell option was not being used for messages that were frozen on
2577 arrival, either by an ACL or by local_scan().
2579 7. Added the smtp_incomplete_transaction log selector.
2581 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2582 was accepting AUTH without a new EHLO.
2584 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2585 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2588 10. Logging of TCP/IP connections (when configured) now happens in the main
2589 daemon process instead of the child process, so that the TCP/IP connection
2590 count is more accurate (but it can never be perfect).
2592 11. The use of "drop" in a nested ACL was not being handled correctly in the
2593 outer ACL. Now, if condition failure induced by the nested "drop" causes
2594 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2595 or "require"), the connection is dropped.
2597 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2598 that yield "discard" can now be used with an "accept" or a "discard" verb,
2599 but an error is generated for any others (because I can't see a useful way
2600 to define what should happen).
2602 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2603 are now processed in the same way as lines in the Exim configuration file.
2604 In particular, continuation lines are supported.
2606 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2608 15. Added -ti meaning -t -i.
2610 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2611 domains, and warn by logging if others are found.
2613 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2614 returned from GIFCONF when Exim is trying to find the list of interfaces on
2615 a host. The code in os.c has been modified to copy each ifreq to an aligned
2616 structure in all cases.
2618 Also, in some cases, the returned ifreq's were being copied to a 'struct
2619 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2620 means the last couple of bytes of an IPv6 address could be chopped if the
2621 ifreq contained only a normal sockaddr (14 bytes storage).
2623 18. Named domain lists were not supported in the hosts_treat_as_local option.
2624 An entry such as +xxxx was not recognized, and was treated as a literal
2627 19. Ensure that header lines added by a DATA ACL are included in the reject log
2628 if the ACL subsequently rejects the message.
2630 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2631 MD5 (which is deprecated).
2633 21. When testing a filter file using -bf, Exim was writing a message when it
2634 took the sender from a "From " line in the message, but it was not doing so
2635 when it took $return_path from a Return-Path: header line. It now does.
2637 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2638 with a valid header line field name (a series of printing characters
2639 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2641 23. Changed "disc" in the source to "disk" to conform to the documentation and
2642 the book and for uniformity.
2644 24. Ignore Sendmail's -Ooption=value command line item.
2646 25. When execve() failed while trying to run a command in a pipe transport,
2647 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2648 could be confused with a return value of 69 from the command itself. This
2649 has been changed to 127, the value the shell returns if it is asked to run
2650 a non-existent command. The wording for the related log line suggests a
2651 non-existent command as the problem.
2653 26. If received_header_text expands to an empty string, do not add a Received:
2654 header line to the message. (Well, it adds a token one on the spool, but
2655 marks it "old" so that it doesn't get used or transmitted.)
2657 27. Installed eximstats 1.28 (addition of -nt option).
2659 28. There was no check for failure on the call to getsockname() in the daemon
2660 code. This can fail if there is a shortage of resources on the system, with
2661 ENOMEM, for example. A temporary error is now given on failure.
2663 29. Contrary to the C standard, it seems that in some environments, the
2664 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2665 program. Exim now does this explicitly; it affects the formatting of
2666 timestamps using strftime().
2668 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2669 complaints. I've now initialized all the variables, to avoid this.
2671 32. Header lines added by a system filter were not being "seen" during
2672 transport-time rewrites.
2674 33. The info_callback() function passed to OpenSSL is set up with type void
2675 (*)(SSL *, int, int), as described somewhere. However, when calling the
2676 function (actually a macro) that sets it up, the type void(*)() is
2677 expected. I've put in a cast to prevent warnings from picky compilers.
2679 34. If a DNS black list lookup found a CNAME record, but there were no A
2680 records associated with the domain it pointed at, Exim crashed.
2682 35. If a DNS black list lookup returned more than one A record, Exim ignored
2683 all but the first. It now scans all returned addresses if a particular IP
2684 value is being sought. In this situation, the contents of the
2685 $dnslist_value variable are a list of all the addresses, separated by a
2688 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2689 to accept a host name and all its aliases if the forward lookup for any of
2690 them yielded the IP address of the incoming connection. Now it accepts only
2691 those names whose forward lookup yields the correct IP address. Any other
2692 names are discarded. This closes a loophole whereby a rogue DNS
2693 administrator could create reverse DNS records to break through a
2694 wildcarded host restriction in an ACL.
2696 37. If a user filter or a system filter that ran in a subprocess used any of
2697 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2698 the wrong values were passed to the pipe command ($thisaddress had the
2699 value of $0, $0 had the value of $1, etc). This bug was introduced by
2700 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2702 38. Improved the line breaking for long SMTP error messages from ACLs.
2703 Previously, if there was no break point between 40 and 75 characters, Exim
2704 left the rest of the message alone. Two changes have been made: (a) I've
2705 reduced the minimum length to 35 characters; (b) if it can't find a break
2706 point between 35 and 75 characters, it looks ahead and uses the first one
2707 that it finds. This may give the occasional overlong line, but at least the
2708 remaining text gets split now.
2710 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2711 file descriptors might be low, and that setting 1000 would always raise it.
2712 It turns out that in some environments, the limit is already over 1000 and
2713 that lowering it causes trouble. So now Exim takes care not to decrease it.
2715 40. When delivering a message, the value of $return_path is set to $sender_
2716 address at the start of routing (routers may change the value). By an
2717 oversight, this default was not being set up when an address was tested by
2718 -bt or -bv, which affected the outcome if any router or filter referred to
2721 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2722 log only when "message" or "log_message" are set. However, if one of the
2723 conditions was an address verification, or a call to a nested ACL, the
2724 messages generated by the underlying test were being passed through. This
2725 no longer happens. The underlying message is available in $acl_verify_
2726 message for both "message" and "log_message" expansions, so it can be
2727 passed through if needed.
2729 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2730 new expansion $bh_ to give the encoded byte string without charset
2731 translation. Translation happens only if iconv() is available; HAVE_ICONV
2732 indicates this at build time. HEADERS_CHARSET gives the charset to
2733 translate to; headers_charset can change it in the configuration, and
2734 "headers charset" can change it in an individual filter file.
2736 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2737 that creates RFC 2047 encoded "words" labels them as that charset instead
2738 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2739 expansion operator; (ii) when Exim creates a From: line for a local
2740 message; (iii) when a header line is rewritten to include a "phrase" part.
2742 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2743 buggy, causing it to skip the first lines of messages whose message ID
2744 ended in 'D'. This would not have bitten before Exim release 4.14, because
2745 message IDs were unlikely to end in 'D' before then. The effect was to have
2746 incorrect size information for certain domains.
2748 45. #include "config.h" was missing at the start of the crypt16.c module. This
2749 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2752 46. If there was a timeout during a "random" callout check, Exim treated it as
2753 a failure of the random address, and carried on sending RSET and the real
2754 address. If the delay was just some slowness somewhere, the response to the
2755 original RCPT would be taken as a response to RSET and so on, causing
2756 mayhem of various kinds.
2758 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2759 when I implemented it. It didn't allow for the fact that some option values
2760 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2761 the right test for positive values.
2763 48. Domain names in DNS records are case-independent. Exim always looks them up
2764 in lower case. Some resolvers return domain names in exactly the case they
2765 appear in the zone file, that is, they may contain uppercase letters. Not
2766 all resolvers do this - some return always lower case. Exim was treating a
2767 change of case by a resolver as a change of domain, similar to a widening
2768 of a domain abbreviation. This triggered its re-routing code and so it was
2769 trying to route what was effectively the same domain again. This normally
2770 caused routing to fail (because the router wouldn't handle the domain
2771 twice). Now Exim checks for this case specially, and just changes the
2772 casing of the domain that it ultimately uses when it transmits the message
2775 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2778 50. If a filter generated a file delivery with a non-absolute name (possible if
2779 no home directory exists for the router), the forbid_file option was not
2782 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2783 the existing equality matching.
2785 52. Exim was using ints instead of ino_t variables in some places where it was
2786 dealing with inode numbers.
2788 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2789 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2790 called TMPDIR, and if it finds it is different, it changes its value.
2792 54. The smtp_printf() function is now made available to local_scan() so
2793 additional output lines can be written before returning. There is also an
2794 smtp_fflush() function to enable the detection of a dropped connection.
2795 The variables smtp_input and smtp_batched_input are exported to
2798 55. Changed the default runtime configuration: the message "Unknown user"
2799 has been removed from the ACL, and instead placed on the localuser router,
2800 using the cannot_route_message feature. This means that any verification
2801 failures that generate their own messages won't get overridden. Similarly,
2802 the "Unrouteable address" message that was in the ACL for unverifiable
2803 relay addresses has also been removed.
2805 56. Added hosts_avoid_esmtp to the smtp transport.
2807 57. The exicyclog script was not checking for the esoteric option
2808 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2809 will work only if exicyclog is run under the appropriate euid.
2811 58. Following a discussion on the list, the rules by which Exim recognises line
2812 endings on incoming messages have been changed. The -dropcr and drop_cr
2813 options are now no-ops, retained only for backwards compatibility. The
2814 following line terminators are recognized: LF CRLF CR. However, special
2815 processing applies to CR:
2817 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2818 nor a local message in the state where . is a terminator.
2820 (ii) If a bare CR is encountered in a header line, an extra space is added
2821 after the line terminator so as not to end the header. The reasoning
2822 behind this is that bare CRs in header lines are most likely either
2823 to be mistakes, or people trying to play silly games.
2825 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2826 was being incorrectly given as 18 bytes larger than it should have been.
2827 This is a VOB (very old bug).
2829 60. This may never have affected anything current, but just in case it has:
2830 When the local host is found other than at the start of a list of hosts,
2831 the local host, those with the same MX, and any that follow, are discarded.
2832 When the list in question was part of a longer list of hosts, the following
2833 hosts (not currently being processed) were also being discarded. This no
2834 longer happens. I'm not sure if this situation could ever has previously
2837 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2840 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2841 header. This is something that is recommended in a new Internet Draft, and
2842 is something that is documented as being done by Sendmail. There are two
2843 possible values. For messages generated by the autoreply transport, Exim
2846 Auto-Submitted: auto-replied
2848 whereas for all other generated messages (e.g. bounces) it adds
2850 Auto-Submitted: auto-generated
2852 63. The "personal" condition in filters now includes a test for the
2853 Auto-Submitted: header. If it contains the string "auto-" the message it
2854 not considered personal.
2856 64. Added rcpt_include_affixes as a generic transport option.
2858 65. Added queue_only_override (default true).
2860 66. Added the syslog_duplication option.
2862 67. If what should have been the first header line of a message consisted of
2863 a space followed by a colon, Exim was mis-interpreting it as a header line.
2864 It isn't of course - it is syntactically invalid and should therefore be
2865 treated as the start of the message body. The misbehaviour could have
2866 caused a number of strange effects, including loss of data in subsequent
2867 header lines, and spool format errors.
2869 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2870 client host had authenticated. This control can now be exercised by an ACL
2871 for more flexibility.
2873 69. By default, callouts do not happen when testing with -bh. There is now a
2874 variant, -bhc, which does actually run the callout code, including
2875 consulting and updating the callout cache.
2877 70. Added support for saslauthd authentication, courtesy of Alexander
2880 71. If statvfs() failed on the spool or log directories while checking their
2881 size for availability, Exim confusingly gave the error "space shortage".
2882 Furthermore, in debugging mode it crashed with a floating point exception.
2883 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2884 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2885 serious problem, Exim now writes to the main and panic logs when this
2886 happens, with details of the failure. It then refuses to accept the
2887 incoming message, giving the message "spool directory problem" or "log
2888 directory problem" with a 421 code for SMTP messages.
2890 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2891 0, 1, and 2 exist, because some OS complain for execs without them (see
2892 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2893 descriptors. However, the code omitted to check that the open succeeded,
2894 causing mysterious errors if for some reason the permissions on /dev/null
2895 got screwed. Now Exim writes a message to the main and panic logs, and
2896 bombs out if it can't open /dev/null.
2898 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2899 interact so that it is all more flexible. It is supposed to remain
2900 backwards compatible. Also added extra_local_interfaces.
2902 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2903 to bomb out with an assertion failure - to the client this appears as a
2904 connection drop. This problem occurs in the part of the code that was taken
2905 from the Samba project. Fortunately, the assertion is in a very simple
2906 function, so I have fixed this by reproducing the function inline in the
2907 one place where it is called, and arranging for authentication to fail
2908 instead of killing the process with assert().
2910 75. The SPA client code was not working when the server requested OEM rather
2911 than Unicode encoding.
2913 76. Added code to make require_files with a specific uid setting more usable in
2914 the case where statting the file as root fails - usually a non-root-mounted
2915 NFS file system. When this happens and the failure is EACCES, Exim now
2916 forks a subprocess and does the per-uid checking as the relevant uid.
2918 77. Added process_log_path.
2920 78. If log_file_path was not explicitly set, a setting of check_log_space or
2921 check_log_inodes was ignored.
2923 79. If a space check for the spool or log partitions fails, the incident is now
2924 logged. Of course, in the latter case the data may get lost...
2926 80. Added the %p formatting code to string_format() so that it can be used to
2927 print addresses in debug_print(). Adjusted all the address printing in the
2928 debugging in store.c to use %p rather than %d.
2930 81. There was a concern that a line of code in smtp_in.c could overflow a
2931 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2932 initially expressed, the concern was not well-founded, because trailing
2933 spaces are removed early. However, if the trailing spaces were followed by
2934 a NULL, they did not get removed, so the overflow was possible. Two fixes
2937 (a) I re-wrote the offending code in a cleaner fashion.
2938 (b) If an incoming SMTP command contains a NULL character, it is rejected
2941 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2942 runs initgroups(), so that if the Exim user is in any additional groups,
2943 they will be used during message reception.
2949 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob