1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.143 2005/05/24 10:57:10 ph10 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.
40 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
42 TK/03 Merged latest SRS patch from Miles Wilton.
44 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
45 with the definition in sysexits.h (which is #included earlier).
46 Fortunately, Exim does not actually use EX_OK. The code used to try to
47 preserve the sysexits.h value, by assumimg that macro definitions were
48 scanned for macro replacements. I have been disabused of this notion,
49 so now the code just undefines EX_OK before #including unistd.h.
55 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
56 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
58 TK/02 Fix ACL "control" statement not being available in MIME ACL.
60 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
62 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
63 to test Sieve filters that use "vacation".
65 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
66 that changes the way the GnuTLS parameters are stored in the cache file.
67 The new format can be generated externally. For backward compatibility,
68 if the data in the cache doesn't make sense, Exim assumes it has read an
69 old-format file, and it generates new data and writes a new file. This
70 means that you can't go back to an older release without removing the
73 PH/03 A redirect router that has both "unseen" and "one_time" set does not
74 work if there are any delivery delays because "one_time" forces the
75 parent to be marked "delivered", so its unseen clone is never tried
76 again. For this reason, Exim now forbids the simultaneous setting of
79 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
80 redirected to themselves ("homonym" addresses). Read the long ChangeLog
81 entry if you want to know the details. The fix, however, neglected to
82 consider the case when local delivery batching is involved. The test for
83 "previously delivered" was not happening when checking to see if an
84 address could be batched with a previous (undelivered) one; under
85 certain circumstances this could lead to multiple deliveries to the same
88 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
89 in its include files, and this causes problems building Exim.
91 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
92 header_syntax) but Exim was just ignoring anything given after a slash.
93 In particular, this caused confusion with an attempt to use "verify =
94 reverse_host_lookup/defer_ok". An error is now given when options are
95 supplied for verify items that do not have them. (Maybe reverse_host_
96 lookup should have a defer_ok option, but that's a different point.)
98 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
99 defined by RFC 821) to 2048, because there were problems with some AUTH
100 commands, and RFC 1869 says the size should be increased for extended
101 SMTP commands that take arguments.
103 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
106 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
107 "unknown" error; now it says that the functionality isn't in the binary.
109 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
110 an address' error message when a string expansion fails (syntax or
111 whatever). Otherwise the password may appear in the log. Following change
112 PH/42 below, there is no longer a chance of it appearing in a bounce
115 PH/11 Installed exipick version 20050225.0 from John Jetmore.
117 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
118 of its addresses was ever tried. (Bugzilla bug #2.)
120 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
121 the result incorrectly in the debug output. (It correctly added a newline
122 to what was transported.)
124 TF/01 Added $received_time.
126 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
127 commented out examples of how to interface to a virus scanner and to
128 SpamAssassin. Also added commented examples of av_scanner and
129 spamd_address settings.
131 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
132 and controls are allowed in which ACLs. There were a couple of minor
133 errors. Some of the entries in the conditions table (which is a table of
134 where they are NOT allowed) were getting very unwieldy; rewrote them as a
135 negation of where the condition IS allowed.
137 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
139 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
140 header file does not have a version number, so I've had to invent a new
141 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
142 API. The code is untested by me (my Linux distribution still has 0.3.2 of
143 radiusclient), but it was contributed by a Radius user.
145 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
146 files or directories, for OpenSSL.
148 PH/19 When an Exim process that is running as root has to create an Exim log
149 file, it does so in a subprocess that runs as exim:exim so as to get the
150 ownership right at creation (otherwise, other Exim processes might see
151 the file with the wrong ownership). There was no test for failure of this
152 fork() call, which would lead to the process getting stuck as it waited
153 for a non-existent subprocess. Forks do occasionally fail when resources
154 run out. I reviewed all the other calls to fork(); they all seem to check
157 PH/20 When checking for unexpected SMTP input at connect time (before writing
158 the banner), Exim was not dealing correctly with a non-positive return
159 from the read() function. If the client had disconnected by this time,
160 the result was a log entry for a synchronization error with an empty
161 string after "input=" when read() returned zero. If read() returned -1
162 (an event I could not check), uninitialized data bytes were printed.
163 There were reports of junk text (parts of files, etc) appearing after
166 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
168 PH/22 Added support for macro redefinition, and (re)definition in between
169 driver and ACL definitions.
171 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
172 forgetting to use the resulting value; it was using the unexpanded value.
174 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
175 hadn't been configured. The fix is from Juergen Kreileder, who
176 understands it better than I do:
178 "Here's what I see happening with three configured cyrus_sasl
179 authenticators configured (plain, login, cram-md5):
181 On startup auth_cyrus_sasl_init() gets called for each of these.
182 This means three calls to sasl_listmech() without a specified mech_list.
183 => SASL tests which mechs of all available mechs actually work
184 => three warnings about OTP not working
185 => the returned list contains: plain, login, cram-md5, digest-md5, ...
187 With the patch, sasl_listmech() also gets called three times. But now
188 SASL's mech_list option is set to the server_mech specified in the the
189 authenticator. Or in other words, the answer from sasl_listmech()
190 gets limited to just the mech you're testing for (which is different
192 => the return list contains just 'plain' or 'login', 'cram-md5' or
193 nothing depending on the value of ob->server_mech.
195 I've just tested the patch: Authentication still works fine,
196 unavailable mechs specified in the exim configuration are still
197 caught, and the auth.log warnings about OTP are gone."
199 PH/25 When debugging is enabled, the contents of the command line are added
200 to the debugging output, even when log_selector=+arguments is not
203 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
204 answer is "GNU", and only if the return is "GNU/something" is the answer
207 PH/27 $acl_verify_message is now set immediately after the failure of a
208 verification in an ACL, and so is available in subsequent modifiers. In
209 particular, the message can be preserved by coding like this:
211 warn !verify = sender
212 set acl_m0 = $acl_verify_message
214 Previously, $acl_verify_message was set only while expanding "message"
215 and "log_message" when a very denied access.
217 PH/28 Modified OS/os.c-Linux with
219 -#ifndef OS_LOAD_AVERAGE
220 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
222 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
223 nomenclature these days.)
225 PH/29 Installed patch from the Sieve maintainer that adds the options
226 sieve_useraddress and sieve_subaddress to the redirect router.
228 PH/30 In these circumstances:
229 . Two addresses routed to the same list of hosts;
230 . First host does not offer TLS;
231 . First host accepts first address;
232 . First host gives temporary error to second address;
233 . Second host offers TLS and a TLS session is established;
234 . Second host accepts second address.
235 Exim incorrectly logged both deliveries with the TLS parameters (cipher
236 and peerdn, if requested) that were in fact used only for the second
239 PH/31 When doing a callout as part of verifying an address, Exim was not paying
240 attention to any local part prefix or suffix that was matched by the
241 router that accepted the address. It now behaves in the same way as it
242 does for delivery: the affixes are removed from the local part unless
243 rcpt_include_affixes is set on the transport.
245 PH/32 Add the sender address, as F=<...>, to the log line when logging a
246 timeout during the DATA phase of an incoming message.
248 PH/33 Sieve envelope tests were broken for match types other than :is. I have
249 applied a patch sanctioned by the Sieve maintainer.
251 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
252 the uid or gid is negative. A case of a negative gid caused this to be
253 noticed. The fix allows for either to be negative.
255 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
256 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
259 PH/36 The change PH/12 above was broken. Fixed it.
261 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
262 the grounds that routing the same address twice would always produce the
263 same answer. This might have been true once, but it is certainly no
264 longer true now. Routing a child address may depend on the previous
265 routing that produced that child. Some complicated redirection strategies
266 went wrong when messages had multiple recipients, and made Exim's
267 behaviour dependent on the order in which the addresses were given.
269 I have moved the duplicate checking until after the routing is complete.
270 Exim scans the addresses that are assigned to local and remote
271 transports, and removes any duplicates. This means that more work will be
272 done, as duplicates will always all be routed, but duplicates are
273 presumably rare, so I don't expect this is of any significance.
275 For deliveries to pipes, files, and autoreplies, the duplicate checking
276 still happens during the routing process, since they are not going to be
279 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
280 It corrects a timeout issue with spamd. This is Ian's comment: "The
281 background is that sometimes spamd either never reads data from a
282 connection it has accepted, or it never writes response data. The exiscan
283 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
284 blindly assumes that writes won't block so it may never time out."
286 PH/39 Allow G after quota size as well as K and M.
288 PH/40 The value set for $authenticated_id in an authenticator may not contain
289 binary zeroes or newlines because the value is written to log lines and
290 to spool files. There was no check on this. Now the value is run through
291 the string_printing() function so that such characters are converted to
292 printable escape sequences.
294 PH/41 $message_linecount is a new variable that contains the total number of
295 lines in the message. Compare $body_linecount, which is the count for the
298 PH/42 Exim no longer gives details of delivery errors for specific addresses in
299 bounce and delay warning messages, except in certain special cases, which
302 (a) An SMTP error message from a remote host;
303 (b) A message specified in a :fail: redirection;
304 (c) A message specified in a "fail" command in a system filter;
305 (d) A message specified in a FAIL return from the queryprogram router;
306 (e) A message specified by the cannot_route_message router option.
308 In these cases only, Exim does include the error details in bounce and
309 warning messages. There are also a few cases where bland messages such
310 as "unrouteable address" or "local delivery error" are given.
312 PH/43 $value is now also set for the "else" part of a ${run expansion.
314 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
315 being worked on, but at least Exim now implements the latest version to
318 PH/45 In a pipe transport, although a timeout while waiting for the pipe
319 process to complete was treated as a delivery failure, a timeout while
320 writing the message to the pipe was logged, but erroneously treated as a
321 successful delivery. Such timeouts include transport filter timeouts. For
322 consistency with the overall process timeout, these timeouts are now
323 treated as errors, giving rise to delivery failures by default. However,
324 there is now a new Boolean option for the pipe transport called
325 timeout_defer, which, if set TRUE, converts the failures into defers for
326 both kinds of timeout. A transport filter timeout is now identified in
329 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
330 systems where "make" and "gmake" are different, calling "gmake" at top
331 level broke things. I've arranged for the value of $(MAKE) to be passed
332 from the Makefile to this script so that it can call the same version of
336 A note about Exim versions 4.44 and 4.50
337 ----------------------------------------
339 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
340 changes of various kinds. As a consequence, a big documentation update was
341 needed. This delayed the release for rather longer than seemed good, especially
342 in the light of a couple of (minor) security issues. Therefore, the changes
343 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
344 release. So 4.44 and 4.50 are in effect two different branches that both start
347 I have left the 4.50 change log unchanged; it contains all the changes since
348 4.43. The change log for 4.44 is below; many of its items are identical to
349 those for 4.50. This seems to be the most sensible way to preserve the
350 historical information.
356 1. Minor wording change to the doc/README.SIEVE file.
358 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
359 computation of the current number of files was incorrect.
361 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
362 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
365 4. Give more explanation in the error message when the command for a transport
366 filter fails to execute.
368 5. There are several places where Exim runs a non-Exim command in a
369 subprocess. The SIGUSR1 signal should be disabled for these processes. This
370 was being done only for the command run by the queryprogram router. It is
371 now done for all such subprocesses. The other cases are: ${run, transport
372 filters, and the commands run by the lmtp and pipe transports.
374 6. Added CONFIGURE_GROUP build-time option.
376 7. Some older OS have a limit of 256 on the maximum number of file
377 descriptors. Exim was using setrlimit() to set 1000 as a large value
378 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
379 systems. I've change it so that if it can't get 1000, it tries for 256.
381 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
382 was an oversight, and furthermore, ever since the addition of extra
383 controls (e.g. 4.43/32), the checks on when to allow different forms of
384 "control" were broken. There should now be diagnostics for all cases when a
385 control that does not make sense is encountered.
387 9. Added the /retain_sender option to "control=submission".
389 10. $recipients is now available in the predata ACL (oversight).
391 11. Tidy the search cache before the fork to do a delivery from a message
392 received from the command line. Otherwise the child will trigger a lookup
393 failure and thereby defer the delivery if it tries to use (for example) a
394 cached ldap connection that the parent has called unbind on.
396 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
397 of $address_data from the recipient verification was clobbered by the
400 13. The value of address_data from a sender verification is now available in
401 $sender_address_data in subsequent conditions in the ACL statement.
403 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
405 15. Added a new option "connect=<time>" to callout options, to set a different
408 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
409 was its contents. (It was OK if the option was not defined at all.)
411 17. A "Completed" log line is now written for messages that are removed from
412 the spool by the -Mrm option.
414 18. New variables $sender_verify_failure and $recipient_verify_failure contain
415 information about exactly what failed.
417 19. Added -dd to debug only the daemon process.
419 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
420 handles timeouts, both on the server side and network timeouts. Renamed the
421 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
423 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
424 It is now set to "smtps".
426 22. $host_address is now set to the target address during the checking of
429 23. When checking ignore_target_hosts for an ipliteral router, no host name was
430 being passed; this would have caused $sender_host_name to have been used if
431 matching the list had actually called for a host name (not very likely,
432 since this list is usually IP addresses). A host name is now passed as
435 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
436 code that specifies a non-restarting handler (typically sigaction() in
437 modern systems) in an attempt to fix a rare and obscure crash bug.
439 25. Narrowed the window for a race in the daemon that could cause it to ignore
440 SIGCHLD signals. This is not a major problem, because they are used only to
441 wake it up if nothing else does.
443 26. A malformed maildirsize file could cause Exim to calculate negative values
444 for the mailbox size or file count. Odd effects could occur as a result.
445 The maildirsize information is now recalculated if the size or filecount
448 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
449 support for a long time. Removed HAVE_SYS_VFS_H.
451 28. Installed the latest version of exipick from John Jetmore.
453 29. In an address list, if the pattern was not a regular expression, an empty
454 subject address (from a bounce message) matched only if the pattern was an
455 empty string. Non-empty patterns were not even tested. This was the wrong
456 because it is perfectly reasonable to use an empty address as part of a
457 database query. An empty address is now tested by patterns that are
458 lookups. However, all the other forms of pattern expect the subject to
459 contain a local part and a domain, and therefore, for them, an empty
460 address still always fails if the pattern is not itself empty.
462 30. Exim went into a mad DNS loop when attempting to do a callout where the
463 host was specified on an smtp transport, and looking it up yielded more
466 31. Re-factored the code for checking spool and log partition space into a
467 function that finds that data and another that does the check. The former
468 is then used to implement four new variables: $spool_space, $log_space,
469 $spool_inodes, and $log_inodes.
471 32. The RFC2047 encoding function was originally intended for short strings
472 such as real names; it was not keeping to the 75-character limit for
473 encoded words that the RFC imposes. It now respects the limit, and
474 generates multiple encoded words if necessary. To be on the safe side, I
475 have increased the buffer size for the ${rfc2047: expansion operator from
478 33. It is now permitted to omit both strings after an "if" condition; if the
479 condition is true, the result is "true". As before, when the second string
480 is omitted, a false condition yields an empty string. This makes it less
481 cumbersome to write custom ACL and router conditions.
483 34. Failure to deliver a bounce message always caused it to be frozen, even if
484 there was an errors_to setting on the router. The errors_to setting is now
487 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
488 canonical form (fully expanded) before being placed in
489 $sender_host_address.
491 36. The table in the code that translates DNS record types into text (T_A to
492 "A" for instance) was missing entries for NS and CNAME. It is just possible
493 that this could have caused confusion if both these types were looked up
494 for the same domain, because the text type is used as part of Exim's
495 per-process caching. But the chance of anyone hitting this buglet seems
498 37. The dnsdb lookup has been extended in a number of ways.
500 (1) There is a new type, "zns", which walks up the domain tree until it
501 finds some nameserver records. It should be used with care.
503 (2) There is a new type, "mxh", which is like "mx" except that it returns
504 just the host names, not the priorities.
506 (3) It is now possible to give a list of domains (or IP addresses) to be
507 looked up. The behaviour when one of the lookups defers can be
508 controlled by a keyword.
510 (4) It is now possible to specify the separator character for use when
511 multiple records are returned.
513 38. The dnslists ACL condition has been extended: it is now possible to supply
514 a list of IP addresses and/or domains to be looked up in a particular DNS
517 39. Added log_selector=+queue_time_overall.
519 40. When running the queue in the test harness, wait just a tad after forking a
520 delivery process, to get repeatability of debugging output.
522 41. Include certificate and key file names in error message when GnuTLS fails
523 to set them up, because the GnuTLS error message doesn't include the name
524 of the failing file when there is a problem reading it.
526 42. Allow both -bf and -bF in the same test run.
528 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
530 44. The "Exiscan patch" is now merged into the mainline Exim source.
532 45. Sometimes the final signoff response after QUIT could fail to get
533 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
534 < 0 before doing a fflush(). This bug looks as though it goes back to the
535 introduction of TLS in release 3.20, but "sometimes" must have been rare
536 because the tests only now provoked it.
538 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
539 (this can affect the format of dates).
541 47. exim_tidydb, when checking for the continued existence of a message for
542 which it has found a message-specific retry record, was not finding
543 messages that were in split spool directories. Consequently, it was
544 deleting retry records that should have stayed in existence.
546 48. Steve fixed some bugs in eximstats.
548 49. The SPA authentication driver was not abandoning authentication and moving
549 on to the next authenticator when an expansion was forced to fail,
550 contradicting the general specification for all authenticators. Instead it
551 was generating a temporary error. It now behaves as specified.
553 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
554 (the order specifies the preference for clients). The order is now AES256,
555 AES128, 3DES, ARCFOUR128.
557 51. Small patch to Sieve code - explicitly set From: when generating an
560 52. Exim crashed if a remote delivery caused a very long error message to be
561 recorded - for instance if somebody sent an entire SpamAssassin report back
562 as a large number of 550 error lines. This bug was coincidentally fixed by
563 increasing the size of one of Exim's internal buffers (big_buffer) that
564 happened as part of the Exiscan merge. However, to be on the safe side, I
565 have made the code more robust (and fixed the comments that describe what
568 53. Now that there can be additional text after "Completed" in log lines (if
569 the queue_time_overall log selector is set), a one-byte patch to exigrep
570 was needed to allow it to recognize "Completed" as not the last thing in
573 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
574 patch that reportedly fixes this has been added. I am not expert enough to
575 create a test for it. This is what the patch creator wrote:
577 "I found a little strange behaviour of ldap code when working with
578 Windows 2003 AD Domain, where users was placed in more than one
579 Organization Units. When I tried to give exim partial DN, the exit code
580 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
581 But simultaneously result of request was absolutely normal ldap result,
582 so I produce this patch..."
584 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
585 REFERENCE, so I have modified the code to exclude the patch when that macro
588 55. Some experimental protocols are using DNS PTR records for new purposes. The
589 keys for these records are domain names, not reversed IP addresses. The
590 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
591 leaves it alone. Component reversal etc. now happens only for IP addresses.
594 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
596 57. Double the size of the debug message buffer (to 2048) so that more of very
597 long debug lines gets shown.
599 58. The exicyclog utility now does better if the number of log files to keep
600 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
602 59. Two changes related to the smtp_active_hostname option:
604 (1) $smtp_active_hostname is now available as a variable.
605 (2) The default for smtp_banner uses $smtp_active_hostname instead
606 of $primary_hostname.
608 60. The host_aton() function is supposed to be passed a string that is known
609 to be a valid IP address. However, in the case of IPv6 addresses, it was
610 not checking this. This is a hostage to fortune. Exim now panics and dies
611 if the condition is not met. A case was found where this could be provoked
612 from a dnsdb PTR lookup with an IPv6 address that had more than 8
613 components; fortuitously, this particular loophole had already been fixed
614 by change 4.50/55 above.
616 If there are any other similar loopholes, the new check in host_aton()
617 itself should stop them being exploited. The report I received stated that
618 data on the command line could provoke the exploit when Exim was running as
619 exim, but did not say which command line option was involved. All I could
620 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
624 61. There was a buffer overflow vulnerability in the SPA authentication code
625 (which came originally from the Samba project). I have added a test to the
626 spa_base64_to_bits() function which I hope fixes it.
629 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
630 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
632 63. The daemon start-up calls getloadavg() while still root for those OS that
633 need the first call to be done as root, but it missed one case: when
634 deliver_queue_load_max is set with deliver_drop_privilege. This is
635 necessary for the benefit of the queue runner, because there is no re-exec
636 when deliver_drop_privilege is set.
638 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
641 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
642 was in use, was not putting the data itself into the right store pool;
643 consequently, it could be overwritten for a subsequent message in the same
644 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
647 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
649 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
650 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
651 the function were treating the return as a boolean value, which happened to
652 work because 0=false and not-0=true, but is not correct code.
654 68. The host_aton() function was not handling scoped IPv6 addresses (those
655 with, for example, "%eth0" on the end) correctly.
657 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
658 negated items (that is, ~something) in unsigned ints. Some compilers
659 apparently mutter when there is no cast.
661 70. If an address verification called from an ACL failed, and did not produce a
662 user-specific message (i.e. there was only a "system" message), nothing was
663 put in $acl_verify_message. In this situation, it now puts the system
666 71. Change 4.23/11 added synchronization checking at the start of an SMTP
667 session; change 4.31/43 added the unwanted input to the log line - except
668 that it did not do this in the start of session case. It now does.
670 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
671 This is wrong and can cause the other end to generate a synchronization
672 error if it is another Exim or anything else that does the synchronization
673 check. A QUIT command is no longer sent after a timeout.
675 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
678 74. The defer_ok option of callout verification was not working if it was used
679 when verifying addresses in header lines, that is, for this case:
681 verify = header_sender/callout=defer_ok
683 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
684 those file descriptors could be used for SMTP connections. If anything
685 wrote to stderr (the example that came up was "warn" in embedded Perl), it
686 could be sent to the SMTP client, causing chaos. The daemon now opens
687 stdin, stdout, and stderr to /dev/null when it puts itself into the
690 76. Arrange for output from Perl's "warn" command to be written to Exim's main
691 log by default. The user can override this with suitable Perl magic.
693 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
694 the log message when discard triggers, was not working for the DATA ACL or
695 for the non-SMTP ACL.
697 78. Error message wording change in sieve.c.
699 79. If smtp_accept_max_per_host was set, the number of connections could be
700 restricted to fewer than expected, because the daemon was trying to set up
701 a new connection before checking whether the processes handling previous
702 connections had finished. The check for completed processes is now done
703 earlier. On busy systems, this bug wouldn't be noticed because something
704 else would have woken the daemon, and it would have reaped the completed
707 80. If a message was submitted locally by a user whose login name contained one
708 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
709 It caused a spool format error. I have fixed the spool reading code. A
710 related problem was that the "from" clause in the Received: line became
711 illegal because of the space(s). It is now covered by ${quote_local_part.
713 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
716 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
718 83. Patch from Sieve maintainer to fix unterminated string problem in
721 84. Some minor changes to the Linux configuration files to help with other
722 OS variants using glibc.
724 85. One more patch for Sieve to update vacation handling to latest spec.
727 ----------------------------------------------------
728 See the note above about the 4.44 and 4.50 releases.
729 ----------------------------------------------------
735 1. Change 4.43/35 introduced a bug that caused file counts to be
736 incorrectly computed when quota_filecount was set in an appendfile
739 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
740 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
743 3. Give more explanation in the error message when the command for a transport
744 filter fails to execute.
746 4. There are several places where Exim runs a non-Exim command in a
747 subprocess. The SIGUSR1 signal should be disabled for these processes. This
748 was being done only for the command run by the queryprogram router. It is
749 now done for all such subprocesses. The other cases are: ${run, transport
750 filters, and the commands run by the lmtp and pipe transports.
752 5. Some older OS have a limit of 256 on the maximum number of file
753 descriptors. Exim was using setrlimit() to set 1000 as a large value
754 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
755 systems. I've change it so that if it can't get 1000, it tries for 256.
757 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
758 was an oversight, and furthermore, ever since the addition of extra
759 controls (e.g. 4.43/32), the checks on when to allow different forms of
760 "control" were broken. There should now be diagnostics for all cases when a
761 control that does not make sense is encountered.
763 7. $recipients is now available in the predata ACL (oversight).
765 8. Tidy the search cache before the fork to do a delivery from a message
766 received from the command line. Otherwise the child will trigger a lookup
767 failure and thereby defer the delivery if it tries to use (for example) a
768 cached ldap connection that the parent has called unbind on.
770 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
771 of $address_data from the recipient verification was clobbered by the
774 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
775 was its contents. (It was OK if the option was not defined at all.)
777 11. A "Completed" log line is now written for messages that are removed from
778 the spool by the -Mrm option.
780 12. $host_address is now set to the target address during the checking of
783 13. When checking ignore_target_hosts for an ipliteral router, no host name was
784 being passed; this would have caused $sender_host_name to have been used if
785 matching the list had actually called for a host name (not very likely,
786 since this list is usually IP addresses). A host name is now passed as
789 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
790 code that specifies a non-restarting handler (typically sigaction() in
791 modern systems) in an attempt to fix a rare and obscure crash bug.
793 15. Narrowed the window for a race in the daemon that could cause it to ignore
794 SIGCHLD signals. This is not a major problem, because they are used only to
795 wake it up if nothing else does.
797 16. A malformed maildirsize file could cause Exim to calculate negative values
798 for the mailbox size or file count. Odd effects could occur as a result.
799 The maildirsize information is now recalculated if the size or filecount
802 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
803 support for a long time. Removed HAVE_SYS_VFS_H.
805 18. Updated exipick to current release from John Jetmore.
807 19. Allow an empty sender to be matched against a lookup in an address list.
808 Previously the only cases considered were a regular expression, or an
811 20. Exim went into a mad DNS lookup loop when doing a callout where the
812 host was specified on the transport, if the DNS lookup yielded more than
815 21. The RFC2047 encoding function was originally intended for short strings
816 such as real names; it was not keeping to the 75-character limit for
817 encoded words that the RFC imposes. It now respects the limit, and
818 generates multiple encoded words if necessary. To be on the safe side, I
819 have increased the buffer size for the ${rfc2047: expansion operator from
822 22. Failure to deliver a bounce message always caused it to be frozen, even if
823 there was an errors_to setting on the router. The errors_to setting is now
826 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
827 canonical form (fully expanded) before being placed in
828 $sender_host_address.
830 24. Updated eximstats to version 1.33
832 25. Include certificate and key file names in error message when GnuTLS fails
833 to set them up, because the GnuTLS error message doesn't include the name
834 of the failing file when there is a problem reading it.
836 26. Expand error message when OpenSSL has problems setting up cert/key files.
839 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
840 (this can affect the format of dates).
842 28. exim_tidydb, when checking for the continued existence of a message for
843 which it has found a message-specific retry record, was not finding
844 messages that were in split spool directories. Consequently, it was
845 deleting retry records that should have stayed in existence.
847 29. eximstats updated to version 1.35
848 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
849 1.35 - bugfix such that pie charts by volume are generated correctly
851 30. The SPA authentication driver was not abandoning authentication and moving
852 on to the next authenticator when an expansion was forced to fail,
853 contradicting the general specification for all authenticators. Instead it
854 was generating a temporary error. It now behaves as specified.
856 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
857 (the order specifies the preference for clients). The order is now AES256,
858 AES128, 3DES, ARCFOUR128.
860 31. Small patch to Sieve code - explicitly set From: when generating an
863 32. Exim crashed if a remote delivery caused a very long error message to be
864 recorded - for instance if somebody sent an entire SpamAssassin report back
865 as a large number of 550 error lines. This bug was coincidentally fixed by
866 increasing the size of one of Exim's internal buffers (big_buffer) that
867 happened as part of the Exiscan merge. However, to be on the safe side, I
868 have made the code more robust (and fixed the comments that describe what
871 33. Some experimental protocols are using DNS PTR records for new purposes. The
872 keys for these records are domain names, not reversed IP addresses. The
873 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
874 leaves it alone. Component reversal etc. now happens only for IP addresses.
877 34. The host_aton() function is supposed to be passed a string that is known
878 to be a valid IP address. However, in the case of IPv6 addresses, it was
879 not checking this. This is a hostage to fortune. Exim now panics and dies
880 if the condition is not met. A case was found where this could be provoked
881 from a dnsdb PTR lookup with an IPv6 address that had more than 8
882 components; fortuitously, this particular loophole had already been fixed
883 by change 4.50/55 or 4.44/33 above.
885 If there are any other similar loopholes, the new check in host_aton()
886 itself should stop them being exploited. The report I received stated that
887 data on the command line could provoke the exploit when Exim was running as
888 exim, but did not say which command line option was involved. All I could
889 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
893 35. There was a buffer overflow vulnerability in the SPA authentication code
894 (which came originally from the Samba project). I have added a test to the
895 spa_base64_to_bits() function which I hope fixes it.
898 36. The daemon start-up calls getloadavg() while still root for those OS that
899 need the first call to be done as root, but it missed one case: when
900 deliver_queue_load_max is set with deliver_drop_privilege. This is
901 necessary for the benefit of the queue runner, because there is no re-exec
902 when deliver_drop_privilege is set.
904 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
905 was in use, was not putting the data itself into the right store pool;
906 consequently, it could be overwritten for a subsequent message in the same
907 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
910 38. Sometimes the final signoff response after QUIT could fail to get
911 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
912 < 0 before doing a fflush(). This bug looks as though it goes back to the
913 introduction of TLS in release 3.20, but "sometimes" must have been rare
914 because the tests only now provoked it.
920 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
921 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
922 Now it returns an integer. A number of calls were still expecting a T/F
923 return. Fortuitously, in all cases, the tests worked in OK situations,
924 which is the norm. However, things would have gone wrong on any write
925 failures on the smtp file descriptor. This function is used when sending
926 messages over SMTP and also when doing verify callouts.
928 2. When Exim is called to do synchronous delivery of a locally submitted
929 message (the -odf or -odi options), it no longer closes stderr before doing
932 3. Implemented the mua_wrapper option.
934 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
936 5. Implemented the functions header_remove(), header_testname(),
937 header_add_at_position(), and receive_remove_recipient(), and exported them
940 6. If an ACL "warn" statement specified the addition of headers, Exim already
941 inserted X-ACL-Warn: at the start if there was no header name. However, it
942 was not making this test for the second and subsequent header lines if
943 there were newlines in the string. This meant that an invalid header could
944 be inserted if Exim was badly configured.
946 7. Allow an ACL "warn" statement to add header lines at the start or after all
947 the Received: headers, as well as at the end.
949 8. Added the rcpt_4xx retry error code.
951 9. Added postmaster_mailfrom=xxx to callout verification option.
953 10. Added mailfrom=xxxx to the callout verification option, for verify=
956 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
957 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
959 12. Inserted some casts to stop certain compilers warning when using pointer
960 differences as field lengths or precisions in printf-type calls (mostly
961 affecting debugging statements).
963 13. Added optional readline() support for -be (dynamically loaded).
965 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
966 same clock tick as a message's arrival, so that its received time was the
967 same as the "first fail" time on the retry record, and that message
968 remained on the queue past the ultimate address timeout, every queue runner
969 would try a delivery (because it was past the ultimate address timeout) but
970 after another failure, the ultimate address timeout, which should have then
971 bounced the address, did not kick in. This was a "< instead of <=" error;
972 in most cases the first failure would have been in the next clock tick
973 after the received time, and all would be well.
975 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
976 being recognized when the domain list was tested by the match_domain
977 condition in an expansion string.
979 16. Added the ${str2b64: operator.
981 17. Exim was always calling setrlimit() to set a large limit for the number of
982 processes, without checking whether the existing limit was already
983 adequate. (It did check for the limit on file descriptors.) Furthermore,
984 errors from getrlimit() and setrlimit() were being ignored. Now they are
985 logged to the main and panic logs, but Exim does carry on, to try to do its
986 job under whatever limits there are.
988 18. Imported PCRE 5.0.
990 19. Trivial typo in log message " temporarily refused connection" (the leading
993 20. If the log selector return_path_on_delivery was set and an address was
994 redirected to /dev/null, the delivery process crashed because it assumed
995 that a return path would always be set for a "successful" delivery. In this
996 case, the whole delivery is bypassed as an optimization, and therefore no
999 21. Internal re-arrangement: the function for sending a challenge and reading
1000 a response while authentication was assuming a zero-terminated challenge
1001 string. It's now changed to take a pointer and a length, to allow for
1002 binary data in such strings.
1004 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1006 23. Exim was not respecting finduser_retries when seeking the login of the
1007 uid under which it was called; it was always trying 10 times. (The default
1008 setting of finduser_retries is zero.) Also, it was sleeping after the final
1009 failure, which is pointless.
1011 24. Implemented tls_on_connect_ports.
1013 25. Implemented acl_smtp_predata.
1015 26. If the domain in control=submission is set empty, Exim assumes that the
1016 authenticated id is a complete email address when it generates From: or
1017 Sender: header lines.
1019 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1020 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1021 chown and chgrp in /bin and hostname in /usr/bin.
1023 28. Exim was keeping the "process log" file open after each use, just as it
1024 does for the main log. This opens the possibility of it remaining open for
1025 long periods when the USR1 signal hits a daemon. Occasional processlog
1026 errors were reported, that could have been caused by this. Anyway, it seems
1027 much more sensible not to leave this file open at all, so that is what now
1030 29. The long-running daemon process does not normally write to the log once it
1031 has entered its main loop, and it closes the log before doing so. This is
1032 so that log files can straightforwardly be renamed and moved. However,
1033 there are a couple of unusual error situations where the daemon does write
1034 log entries, and I had neglected to close the log afterwards.
1036 30. The text of an SMTP error response that was received during a remote
1037 delivery was being truncated at 512 bytes. This is too short for some of
1038 the long messages that one sometimes sees. I've increased the limit to
1041 31. It is now possible to make retry rules that apply only when a message has a
1042 specific sender, in particular, an empty sender.
1044 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1045 it possible to be selective about when SMTP synchronization is enforced.
1047 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1049 32. Implemented hosts_connection_nolog.
1051 33. Added an ACL for QUIT.
1053 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1056 35. Added mailbox_size and mailbox_filecount to appendfile.
1058 36. Added control = no_multiline_responses to ACLs.
1060 37. There was a bug in the logic of the code that waits for the clock to tick
1061 in the case where the clock went backwards by a substantial amount such
1062 that the microsecond fraction of "now" was more than the microsecond
1063 fraction of "then" (but the whole seconds number was less).
1065 38. Added support for the libradius Radius client library this is found on
1066 FreeBSD (previously only the radiusclient library was supported).
1072 1. When certain lookups returned multiple values in the form name=value, the
1073 quoting of the values was not always being done properly. Specifically:
1074 (a) If the value started with a double quote, but contained no whitespace,
1076 (b) If the value contained whitespace other than a space character (i.e.
1077 tabs or newlines or carriage returns) it was not quoted.
1078 This fix has been applied to the mysql and pgsql lookups by writing a
1079 separate quoting function and calling it from the lookup code. The fix
1080 should probably also be applied to nisplus, ibase and oracle lookups, but
1081 since I cannot test any of those, I have not disturbed their existing code.
1083 2. A hit in the callout cache for a specific address caused a log line with no
1084 reason for rejecting RCPT. Now it says "Previous (cached) callout
1085 verification failure".
1087 3. There was an off-by-one bug in the queryprogram router. An over-long
1088 return line was truncated at 256 instead of 255 characters, thereby
1089 overflowing its buffer with the terminating zero. As well as fixing this, I
1090 have increased the buffer size to 1024 (and made a note to document this).
1092 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1093 when Exim is waiting for an SMTP response from a remote server, Exim
1094 restarts its select() call on the socket, thereby resetting its timeout.
1095 This is not a problem when such interrupts are rare. Somebody set up a cron
1096 job to run exiwhat every 2 minutes, which is less than the normal select()
1097 timeout (5 or 10 minutes). This meant that the select() timeout never
1098 kicked in because it was always reset. I have fixed this by comparing the
1099 time when an interrupt arrives with the time at the start of the first call
1100 to select(). If more time than the timeout has elapsed, the interrupt is
1101 treated as a timeout.
1103 5. Some internal re-factoring in preparation for the addition of Sieve
1104 extensions (by MH). In particular, the "personal" test is moved to a
1105 separate function, and given an option for scanning Cc: and Bcc: (which is
1106 not set for Exim filters).
1108 6. When Exim created an email address using the login of the caller as the
1109 local part (e.g. when creating a From: or Sender: header line), it was not
1110 quoting the local part when it contained special characters such as @.
1112 7. Installed new OpenBSD configuration files.
1114 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1115 try to make them clearer.
1117 9. Callout options, other than the timeout value, were being ignored when
1118 verifying sender addresses in header lines. For example, when using
1120 verify = header_sender/callout=no_cache
1122 the cache was (incorrectly) being used.
1124 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1125 only the Cygwin environment.
1127 11. When return_path_on_delivery was set as a log selector, if different remote
1128 addresses in the same message used different return paths and parallel
1129 remote delivery occurred, the wrong values would sometimes be logged.
1130 (Whenever a remote delivery process finished, the return path value from
1131 the most recently started remote delivery process was logged.)
1133 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1134 lines when AUTH and/or TLS are in use. This is the "received protocol"
1135 field. Exim used to use "asmtp" for authenticated SMTP, without any
1136 indication (in the protocol name) for TLS use. Now it follows the RFC and
1137 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1138 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1139 names appear in log lines as well as in Received: header lines.
1141 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1142 extensions, and comparison tests, and to fix some bugs.
1144 14. Changes to the "personal" filter test:
1146 (1) The test was buggy in that it was just doing the equivalent of
1147 "contains" tests on header lines. For example, if a user's address was
1148 anne@some.where, the "personal" test would incorrectly be true for
1150 To: susanne@some.where
1152 This test is now done by extracting each address from the header in turn,
1153 and checking the entire address. Other tests that are part of "personal"
1154 are now done using regular expressions (for example, to check local parts
1155 of addresses in From: header lines).
1157 (2) The list of non-personal local parts in From: addresses has been
1158 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1159 taken from the Sieve specification recommendations.
1161 (3) If the message contains any header line starting with "List-" it is
1162 treated as non-personal.
1164 (4) The test for "circular" in the Subject: header line has been removed
1165 because it now seems ill-conceived.
1167 15. Minor typos in src/EDITME comments corrected.
1169 16. Installed latest exipick from John Jetmore.
1171 17. If headers_add on a router specified a text string that was too long for
1172 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1173 of string_sprintf() is now avoided.
1175 18. $message_body_size was not set (it was always zero) when running the DATA
1176 ACL and the local_scan() function.
1178 19. For the "mail" command in an Exim filter, no default was being set for
1179 the once_repeat time, causing a random time value to be used if "once" was
1180 specified. (If the value happened to be <= 0, no repeat happened.) The
1181 default is now 0s, meaning "never repeat". The "vacation" command was OK
1182 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1183 (I found it when inspecting the code).
1185 20. There is now an overall timeout for performing a callout verification. It
1186 defaults to 4 times the callout timeout, which applies to individual SMTP
1187 commands during the callout. The overall timeout applies when there is more
1188 than one host that can be tried. The timeout is checked before trying the
1189 next host. This prevents very long delays if there are a large number of
1190 hosts and all are timing out (e.g. when the network connections are timing
1191 out). The value of the overall timeout can be changed by specifying an
1192 additional sub-option for "callout", called "maxwait". For example:
1194 verify = sender/callout=5s,maxwait=20s
1196 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1197 to the end before writing, but this should make it even safer).
1199 22. Exim was forgetting that it had advertised PIPELINING for the second and
1200 subsequent messages on an SMTP connection. It was also not resetting its
1201 memory on STARTTLS and an internal HELO.
1203 23. When Exim logs an SMTP synchronization error within a session, it now
1204 records whether PIPELINING has been advertised or not.
1206 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1207 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1208 rather than long int.
1210 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1212 26. Added the never_mail option to autoreply.
1218 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1219 crash if the getsockname() call failed; this can happen if a connection is
1220 closed very soon after it is established. The problem was simply in the
1221 order in which certain operations were done, causing Exim to try to write
1222 to the SMTP stream before it had set up the file descriptor. The bug has
1223 been fixed by making things happen in the correct order.
1229 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1230 before the connection was closed, thus losing the rejection response.
1232 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1233 some early Solaris releases, but causes trouble in current releases where
1234 socklen_t is defined.
1236 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1239 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1242 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1243 that releasing the top of it at the end releases what was used for sub-
1244 expansions (unless the block got too big). However, discard this block if
1245 the first thing is a variable or header, so that we can use its block when
1246 it is dynamic (useful for very large $message_headers, for example).
1248 6. Lookups now cache *every* query, not just the most recent. A new, separate
1249 store pool is used for this. It can be recovered when all lookup caches are
1250 flushed. Lookups now release memory at the end of their result strings.
1251 This has involved some general refactoring of the lookup sources.
1253 7. Some code has been added to the store_xxx() functions to reduce the amount
1254 of flapping under certain conditions.
1256 8. log_incoming_interface used to affect only the <= reception log lines. Now
1257 it causes the local interface and port to be added to several more SMTP log
1258 lines, for example "SMTP connection from", and rejection lines.
1260 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1262 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1264 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1265 could be overwritten at the end of the current message (or the start of a
1266 new message if it was set in a HELO ACL). The value is now preserved for
1267 the duration of the SMTP connection.
1269 12. If a transport had a headers_rewrite setting, and a matching header line
1270 contained an unqualified address, that address was qualified, even if it
1271 did not match any rewriting rules. The underlying bug was that the values
1272 of the flags that permit the existence of unqualified sender and recipient
1273 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1274 non-local messages, and by -bnq for local messages) were not being
1275 preserved with the message after it was received.
1277 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1278 "next input=" as part of the text comprising the host identity instead of
1279 the correct text. The code was using the same buffer for two different
1280 strings. However, depending on which order the printing function evaluated
1281 its arguments, the bug did not always show up. Under Linux, for example, my
1282 test suite worked just fine.
1284 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1285 doesn't work with some older versions of Perl. It has been changed to "my",
1286 which in any case is probably the better facility to use.
1288 15. A really picky compiler found some instances of statements for creating
1289 error messages that either had too many or two few arguments for the format
1292 16. The size of the buffer for calls to the DNS resolver has been increased
1293 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1294 for addresses that have a lot of PTR records. This alleviates a problem; it
1295 does not fully solve it.
1297 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1298 buffer now truncates the list and logs the incident, which is the same
1299 action as happens when Exim is looking up a host name and its aliases.
1300 Previously in this situation something unpredictable would happen;
1301 sometimes it was "internal error: store_reset failed".
1303 18. If a server dropped the connection unexpectedly when an Exim client was
1304 using GnuTLS and trying to read a response, the client delivery process
1305 crashed while trying to generate an error log message.
1307 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1308 string, for example:
1310 warn message = H1: something\nH2: something
1312 the text was added as a single header line from Exim's point of view
1313 though it ended up OK in the delivered message. However, searching for the
1314 second and subsequent header lines using $h_h2: did not work. This has been
1315 fixed. Similarly, if a system filter added multiple headers in this way,
1316 the routers could not see them.
1318 20. Expanded the error message when iplsearch is called with an invalid key to
1319 suggest using net-iplsearch in a host list.
1321 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1322 ACLs are no longer actually imposed (and a message to that effect is
1325 22. If a "gecos" field in a passwd entry contained escaped characters, in
1326 particular, if it contained a \" sequence, Exim got it wrong when building
1327 a From: or a Sender: header from that name. A second bug also caused
1328 incorrect handling when an unquoted " was present following a character
1329 that needed quoting.
1331 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1332 was not being matched caselessly.
1334 24. Arranged for all hyphens in the exim.8 source to be escaped with
1337 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1338 properly. Recipient callout cache records were still being keyed to include
1339 the sender, even when use_sender was set false. This led to far more
1340 callouts that were necessary. The sender is no longer included in the key
1341 when use_sender is false.
1343 26. Added "control = submission" modifier to ACLs.
1345 27. Added the ${base62d: operator to decode base 62 numbers.
1347 28. dnsdb lookups can now access SRV records.
1349 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1350 the configuration file.
1352 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1353 (-v) mode. This makes the output for a verbose queue run more intelligible.
1355 31. Added a use_postmaster feature to recipient callouts.
1357 32. Added the $body_zerocount variable, containing the number of binary zero
1358 bytes in the message body.
1360 33. The time of last modification of the "new" subdirectory is now used as the
1361 "mailbox time last read" when there is a quota error for a maildir
1364 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1366 35. Added +ignore_unknown as a special item in host lists.
1368 36. Code for decoding IPv6 addresses in host lists is now included, even if
1369 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1370 address was recognized as an IP address, but was then not correctly decoded
1371 into binary, causing unexpected and incorrect effects when compared with
1378 1. Very minor rewording of debugging text in manualroute to say "list of
1379 hosts" instead of "hostlist".
1381 2. If verify=header_syntax was set, and a header line with an unqualified
1382 address (no domain) and a large number of spaces between the end of the
1383 name and the colon was received, the reception process suffered a buffer
1384 overflow, and (when I tested it) crashed. This was caused by some obsolete
1385 code that should have been removed. The fix is to remove it!
1387 3. When running in the test harness, delay a bit after writing a bounce
1388 message to get a bit more predictability in the log output.
1390 4. Added a call to search_tidyup() just before forking a reception process. In
1391 theory, someone could use a lookup in the expansion of smtp_accept_max_
1392 per_host which, without the tidyup, could leave open a database connection.
1394 5. Added the variables $recipient_data and $sender_data which get set from a
1395 lookup success in an ACL "recipients" or "senders" condition, or a router
1396 "senders" option, similar to $domain_data and $local_part_data.
1398 6. Moved the writing of debug_print from before to after the "senders" test
1401 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1402 problems for message scanning, either using a data ACL, or using
1403 local_scan() because the Received: header was not generated till after they
1404 were called (in order to set the time as the time of reception completion).
1405 I have revised the way this works. The header is now generated after the
1406 body is received, but before the ACL or local_scan() are called. After they
1407 are run, the timestamp in the header is updated.
1413 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1414 before starting a queue runner without re-exec. This happened only when
1415 deliver_drop_privilege was set or when the Exim user was set to root. The
1416 effect of the bug was that timeouts during subsequent deliveries caused
1417 crashes instead of being properly handled. The handler is now left at its
1418 default (and expected) setting.
1420 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1421 message, again when deliver_drop_privilege is set or Exim is run as root.
1422 The bug described in (1) was not present in this case, but the tidying up
1423 of the other signals was missing. I have made the two cases consistent.
1425 3. The ignore_target_hosts setting on a manualroute router was being ignored
1426 for hosts that were looked up using the /MX notation.
1428 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1431 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1432 operated on the sender address. After changing the $sender_address to <>
1433 for the sender address verify, Exim was re-instated it as the original
1434 (before rewriting) address, but remembering that it had rewritten it, so it
1435 wasn't rewriting it again. This bug also had the effect of breaking the
1436 sender address verification caching when the sender address was rewritten.
1438 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1439 This has been changed so that if the ip literal address matches
1440 ignore_target_hosts, the router declines.
1442 7. Added expansion conditions match_domain, match_address, and match_local_
1443 part (NOT match_host).
1445 8. The placeholder for the Received: header didn't have a length field set.
1447 9. Added code to Exim itself and to exim_lock to test for a specific race
1448 condition that could lead to file corruption when using MBX delivery. The
1449 issue is with the lockfile that is created in /tmp. If this file is removed
1450 after a process has opened it but before that process has acquired a lock,
1451 there is the potential for a second process to recreate the file and also
1452 acquire a lock. This could lead to two Exim processes writing to the file
1453 at the same time. The added code performs the same test as UW imapd; it
1454 checks after acquiring the lock that its file descriptor still refers to
1455 the same named file.
1457 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1458 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1461 11. Added the smtp_active_hostname option. If used, this will typically be made
1462 to depend on the incoming interface address. Because $interface_address is
1463 not set up until the daemon has forked a reception process, error responses
1464 that can happen earlier (such as "too many connections") no longer contain
1467 12. If an expansion in a condition on a "warn" statement fails because a lookup
1468 defers, the "warn" statement is abandoned, and the next ACL statement is
1469 processed. Previously this caused the whole ACL to be aborted.
1471 13. Added the iplsearch lookup type.
1473 14. Added ident_timeout as a log selector.
1475 15. Added tls_certificate_verified as a log selector.
1477 16. Added a global option tls_require_ciphers (compare the smtp transport
1478 option of the same name). This controls incoming TLS connections.
1480 17. I finally figured out how to make tls_require_ciphers do a similar thing
1481 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1482 before starting the TLS session.
1484 18. Tabs are now shown as \t in -bP output.
1486 19. If the log selector return_path_on_delivery was set, Exim crashed when
1487 bouncing a message because it had too many Received: header lines.
1489 20. If two routers both had headers_remove settings, and the first one included
1490 a superfluous trailing colon, the final name in the first list and the
1491 first name in the second list were incorrectly joined into one item (with a
1492 colon in the middle).
1498 1. Added -C and -D options to the exinext utility, mainly to make it easier
1499 to include in the automated testing, but these could be helpful when
1500 multiple configurations are in use.
1502 2. The exinext utility was not formatting the output nicely when there was
1503 an alternate port involved in the retry record key, nor when there was a
1504 message id as well (for retries that were specific to a specific message
1505 and a specific host). It was also confused by IPv6 addresses, because of
1506 the additional colons they contain. I have fixed the IPv4 problem, and
1507 patched it up to do a reasonable job for IPv6.
1509 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1510 delivery, the log line now contains "pipelined" if PIPELINING was used.
1512 4. An SMTP transport process used to panic and die if the bind() call to set
1513 an explicit outgoing interface failed. This has been changed; it is now
1514 treated in the same way as a connect() failure.
1516 5. A reference to $sender_host_name in the part of a conditional expansion
1517 that was being skipped was still causing a DNS lookup. This no longer
1520 6. The def: expansion condition was not recognizing references to header lines
1521 that used bh_ and bheader_.
1523 7. Added the _cache feature to named lists.
1525 8. The code for checking quota_filecount in the appendfile transport was
1526 allowing one more file than it should have been.
1528 9. For compatibility with Sendmail, the command line option
1536 and sets the incoming protocol and host name (for trusted callers). The
1537 host name and its colon can be omitted when only the protocol is to be set.
1538 Note the Exim already has two private options, -pd and -ps, that refer to
1539 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1540 "s", but I don't think that's a major issue.
1542 10. A number of refactoring changes to the code, none of which should affect
1545 (a) The number of logging options was getting close to filling up the
1546 32-bit word that was used as a bit map. I have split them into two classes:
1547 those that are passed in the argument to log_write(), and those that are
1548 only ever tested independently outside of that function. These are now in
1549 separate 32-bit words, so there is plenty of room for expansion again.
1550 There is no change in the user interface or the logging behaviour.
1552 (b) When building, for example, log lines, the code previously used a
1553 macro that called string_cat() twice, in order to add two strings. This is
1554 not really sufficiently general. Furthermore, there was one instance where
1555 it was actually wrong because one of the argument was used twice, and in
1556 one call a function was used. (As it happened, calling the function twice
1557 did not affect the overall behaviour.) The macro has been replaced by a
1558 function that can join an arbitrary number of extra strings onto a growing
1561 (c) The code for expansion conditions now uses a table and a binary chop
1562 instead of a serial search (which was left over from when there were very
1563 few conditions). Also, it now recognizes conditions like "pam" even when
1564 the relevant support is not compiled in: a suitably worded error message is
1565 given if an attempt is made to use such a condition.
1567 11. Added ${time_interval:xxxxx}.
1569 12. A bug was causing one of the ddress fields not to be passed back correctly
1570 from remote delivery subprocesses. The field in question was not being
1571 subsequently used, so this caused to problems in practice.
1573 13. Added new log selectors queue_time and deliver_time.
1575 14. Might have fixed a bug in maildirsizefile handling that threw up
1576 "unexpected character" debug warnings, and recalculated the data
1577 unnecessarily. In any case, I expanded the warning message to give more
1580 15. Added the message "Restricted characters in address" to the statements in
1581 the default ACL that block characters like @ and % in local parts.
1583 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1584 Three changes have been made:
1586 (a) There was a serious bug; a negative response to MAIL caused the whole
1587 recipient domain to be cached as invalid, thereby blocking all messages
1588 to all local parts at the same domain, from all senders. This bug has
1589 been fixed. The domain is no longer cached after a negative response to
1590 MAIL if the sender used is not empty.
1592 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1595 (c) A new callout option, "use_sender" has been added for people who want
1596 the modified behaviour.
1602 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1605 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1606 indeed breaks things for older releases.
1608 3. Added additional logging to the case where there is a problem reading data
1609 from a filter that is running in a subprocess using a pipe, in order to
1610 try to track down a specific problem.
1612 4. Testing facility fudge: when running in the test harness and attempting
1613 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1614 getting "No route to host". Convert this to a timeout.
1616 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1619 6. Some OS don't have socklen_t but use size_t instead. This affects the
1620 fifth argument of getsockopt() amongst other things. This is now
1621 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1622 can be set for individual OS. I have set it for SunOS5, OSF1, and
1623 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1624 some earlier ones do not.
1626 7. Change 4.30/15 was not doing the test caselessly.
1628 8. The standard form for an IPv6 address literal was being rejected by address
1629 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1630 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1631 this, as well as the form without the "IPv6" on the front (but only when
1632 address literals are enabled, of course).
1634 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1636 10. Exim crashed if a message with an empty sender address specified by -f
1637 encountered a router with an errors_to setting. This could be provoked only
1638 by a command such as
1642 where an empty string was supplied; "<>" did not hit this bug.
1644 11. Installed PCRE release 4.5.
1646 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1647 remained set. It is now erased.
1649 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1650 times from message ids (which are base 36 rather than the normal 62).
1652 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1653 were being counted as actual protocol errors, and logged if the log
1654 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1655 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1656 and DATA following a set of rejected RCPTs do not count as protocol errors.
1657 In other words, Exim assumes they were pipelined, though this may not
1658 actually be the case. Of course, in all cases the client gets an
1659 appropriate error code.
1661 15. If a lookup fails in an ACL condition, a message about the failure may
1662 be available; it is used if testing the ACL cannot continue, because most
1663 such messages specify what the cause of the deferral is. However, some
1664 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1665 that caused an old message to be retained and used if a later statement
1666 caused a defer, replacing the real cause of the deferral.
1668 16. If an IP address had so many PTR records that the DNS lookup buffer
1669 was not large enough to hold them, Exim could crash while trying to process
1670 the truncated data. It now detects and logs this case.
1672 17. Further to 4.21/58, another change has been made: if (and only if) the
1673 first line of a message (the first header line) ends with CRLF, a bare LF
1674 in a subsequent header line has a space inserted after it, so as not to
1675 terminate the header.
1677 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1678 unnecessarily, used atoi() instead of strtol(), and didn't check the
1679 termination when getting file sizes from file names by regex.
1681 19. Completely re-implemented the support for maildirsize files, in the light
1682 of a number of problems with the previous contributed implementation
1683 (4.30/29). In particular:
1685 . If the quota is zero, the maildirsize file is maintained, but no quota is
1688 . If the maildir directory does not exist, it is created before any attempt
1689 to write a maildirsize file.
1691 . The quota value in the file is just a cache; if the quota is changed in
1692 the transport, the new value overrides.
1694 . A regular expression is available for excluding directories from the
1697 20. The autoreply transport checks the characters in options that define the
1698 message's headers; it allows continued headers, but it was checking with
1699 isspace() after an embedded newline instead of explicitly looking for a
1702 21. If all the "regular" hosts to which an address was routed had passed their
1703 expiry times, and had not reached their retry times, the address was
1704 bounced, even if fallback hosts were defined. Now Exim should go on to try
1707 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1708 equivalent code in the SMTP transport. Some hosts send humungous responses
1709 to HELO/EHLO, more than 1024 it seems.
1711 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1712 gives compiler warnings in some environments. I've now done it "properly",
1715 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1716 (because of problems with the built-in one) was declared to return uschar *
1717 instead of char *, causing compiler failure.
1719 25. Fixed a file descriptor leak when processing alias/forward files.
1721 26. Fixed a minor format string issue in dbfn.c.
1723 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1725 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1726 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1729 29. When a sender address is verified, it is cached, to save repeating the test
1730 when there is more than one recipient in a message. However, when the
1731 verification involves a callout, it is possible for different callout
1732 options to be set for different recipients. It is too complicated to keep
1733 track of this in the cache, so now Exim always runs a verification when a
1734 callout is required, relying on the callout cache for the optimization.
1735 The overhead is duplication of the address routing, but this should not be
1738 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1739 to be verified with callout=postmaster, and the main callout worked but the
1740 postmaster check failed, the verification correctly failed. However, if a
1741 subsequent RCPT command asked for sender verification *without* the
1742 postmaster check, incorrect caching caused this verification also to fail,
1745 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1746 it was not caching the DNS options (qualify_single, search_parents) that
1747 were used when the lookup failed. A subsequent lookup with different
1748 options therefore always gave the same answer, though there were cases
1749 where it should not have. (Example: a "domains = !$mx_any" option on a
1750 dnslookup router: the "domains" option is always processed without any
1751 widening, but the router might have qualify_single set.) Now Exim uses the
1752 cached value only when the same options are set.
1754 32. Added John Jetmore's "exipick" utility to the distribution.
1756 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1757 than a timeout (e.g. a certificate is required, and is not provided), an
1758 Exim server now closes the connection immediately. Previously it waited for
1759 the client to close - but if the client is SSL, it seems that they each
1760 wait for each other, leading to a delay before one of them times out.
1762 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1763 maintained 0.8.x compatibility because I don't think many are using it, and
1764 it is clearly obsolete.
1766 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1769 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1770 client certificate was expired. A simple patch fixes this, though I don't
1771 understand the full logic of why the verify callback is called multiple
1774 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1775 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1776 which causes problems with some clients (such as the Certicom SSL Plus
1777 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1778 disables the coutermeasure allowing Eudora to connect."
1780 38. Exim was not checking that a write() to a log file succeeded. This could
1781 lead to Bad Things if a log got too big, in particular if it hit a file
1782 size limit. Exim now panics and dies if it cannot write to a log file, just
1783 as it does if it cannot open a log file.
1785 39. Modified OS/Makefile-Linux so that it now contains
1787 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1789 The two -D definitions ensure that Exim is compiled with large file
1790 support, which makes it possible to handle log files that are bigger than
1793 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1794 instance) a domain was checked against a named list that involved a lookup,
1795 causing $domain_data to be set, then another domain was checked against the
1796 same list, then the first domain was re-checked, the value of $domain_data
1797 after the final check could be wrong. In particular, if the second check
1798 failed, it could be set empty. This bug probably also applied to
1801 41. The strip_trailing_dot option was not being applied to the address given
1802 with the -f command-line option.
1804 42. The code for reading a message's header from the spool was incrementing
1805 $received_count, but never initializing it. This meant that the value was
1806 incorrect (doubled) while delivering a message in the same process in which
1807 it was received. In the most common configuration of Exim, this never
1808 happens - a fresh exec is done - but it can happen when
1809 deliver_drop_privilege is set.
1811 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1812 it now includes up to 150 characters of the unexpected data in the log
1815 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1816 and building data strings. The size of both of these buffers was 10 000
1817 bytes - far larger than anybody would *ever* want, thought I. Needless to
1818 say, somebody hit the limit. I have increased the maximum line length to
1819 20 000 and the maximum data length of concatenated lines to 100 000. I have
1820 also fixed two bugs, because there was no checking on these buffers. Tsk,
1821 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1822 buffer is too small.
1824 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1825 lsearch lookups. Now it does.
1827 46. When parsing a route_list item in a manualroute router, a fixed-length
1828 buffer was used for the list of hosts. I made this 1024 bytes long,
1829 thinking that nobody would ever have a list of hosts that long. Wrong.
1830 Somebody had a whole pile of complicated expansion conditions, and the
1831 string was silently truncated, leading to an expansion error. It turns out
1832 that it is easier to change to an unlimited length (owing to other changes
1833 that have happened since this code was originally written) than to build
1834 structure for giving a limitation error. The length of the item that
1835 expands into the list of hosts is now unlimited.
1837 47. The lsearch lookup could not handle data where the length of text line was
1838 more than 4095 characters. Such lines were truncated, leading to shortened
1839 data being returned. It should now handle lines of any length.
1841 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1842 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1845 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1846 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1847 no longer happens in comment lines. A list of replacements is now placed
1848 at the head of all of the source files, except those whose only change is
1849 to replace PERL_COMMAND in the very first #! line.
1851 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1852 messages on the queue, with a bottom-up merge sort, using code contributed
1853 by Michael Haardt. This should make operations like -bp somewhat faster on
1854 large queues. It won't affect queue runners, except when queue_run_in_order
1857 51. Installed eximstats 1.31 in the distribution.
1859 52. Added support for SRV lookups to the dnslookup router.
1861 53. If an ACL referred to $message_body or $message_body_end, the value was not
1862 reset for any messages that followed in the same SMTP session.
1864 54. The store-handling optimization for building very long strings was not
1865 differentiating between the different store pools. I don't think this
1866 actually made any difference in practice, but I've tidied it.
1868 55. While running the routers to verify a sender address, $sender_address
1869 was still set to the sender address. This is wrong, because when routing to
1870 send a bounce to the sender, it would be empty. Therefore, I have changed
1871 it so that, while verifying a sender address, $sender_address is set to <>.
1872 (There is no change to what happens when verifying a recipient address.)
1874 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1875 target A or AAAA records (if not already returned) without resetting the
1876 qualify_single or search_parents options of the DNS resolver. These are
1877 inappropriate in this case because the targets of MX and SRV records must
1878 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1879 target that, when qualified, matched something in the local domain. These
1880 two options are now turned off when doing these lookups.
1882 57. It seems that at least some releases of Reiserfs (which does not have the
1883 concept of a fixed number of inodes) returns zero and not -1 for the
1884 number of available inodes. This interacted badly with check_spool_inodes,
1885 which assumed that -1 was the "no such thing" setting. What I have done is
1886 to check that the total number of inodes is greater than zero before doing
1887 the test of how many are available.
1889 58. When a "warn" ACL statement has a log_message modifier, the message is
1890 remembered, and not repeated. This is to avoid a lot of repetition when a
1891 message has many recipients that cause the same warning to be written.
1892 Howewer, Exim was preserving the list of already written lines for an
1893 entire SMTP session, which doesn't seem right. The memory is now reset if a
1894 new message is started.
1896 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1897 debugging output unless log_rewrite was also set.
1899 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1900 of (int)(handle) when we know that handle contains (void *)(-1).
1902 61. The Exim daemon panic-logs an error return when it closes the incoming
1903 connection. However "connection reset by peer" seems to be common, and
1904 isn't really an error worthy of noting specially, so that particular error
1907 62. When Exim is trying to find all the local interfaces, it used to panic and
1908 die if the ioctl to get the interface flags failed. However, it seems that
1909 on at least one OS (Solaris 9) it is possible to have an interface that is
1910 included in the list of interfaces, but for which you get a failure error
1911 for this call. This happens when the interface is not "plumbed" into a
1912 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1913 failure of the "get flags" call assumes that the interface is down.
1915 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1916 makes life easier for people who are doing arithmetic on fields extracted
1917 from dates, where you often get leading zeros that should not be
1918 interpreted as octal.
1920 64. Added qualify_domain to the redirect router, to override the global
1923 65. If a pathologically long header line contained very many addresses (the
1924 report of this problem mentioned 10 000) and each of them was rewritten,
1925 Exim could use up a very large amount of memory. (It kept on making new
1926 copies of the header line as it rewrote, and never released the old ones.)
1927 At the expense of a bit more processing, the header rewriting function has
1928 been changed so that it no longer eats memory in this way.
1930 66. The generation of the Received: header has been moved from the time that a
1931 message starts to be received, to the time that it finishes. The timestamp
1932 in the Received: header should now be very close to that of the <= log
1933 line. There are two side-effects of this change:
1935 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1936 logged header lines no longer include the local Received: line, because
1937 it has not yet been created. The same applies to a copy of the message
1938 that is returned to a non-SMTP sender when a message is rejected.
1940 (b) When a filter file is tested using -bf, no additional Received: header
1941 is added to the test message. After some thought, I decided that this
1944 This change does not affect the value of $received_for. It is still set
1945 after address rewriting, but before local_scan() is called.
1947 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1949 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1950 gave an unhelpful panic error message, and a defer error. I have managed to
1951 change this behaviour so that it now rejects any supplied certificate,
1952 which seems right, as the list of acceptable certificates is empty.
1954 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1955 gave an unhelpful defer error. I have not managed to make this reject any
1956 supplied certificates, but the error message it gives is "no certificate
1957 supplied", which is not helpful.
1959 70. exigrep's output now also includes lines that are not associated with any
1960 message, but which match the given pattern. Implemented by a patch from
1961 Martin Sluka, which also tidied up the Perl a bit.
1963 71. Recipient callout verification, like sender verification, was using <> in
1964 the MAIL FROM command. This isn't really the right thing, since the actual
1965 sender may affect whether the remote host accepts the recipient or not. I
1966 have changed it to use the actual sender in the callout; this means that
1967 the cache record is now keyed on a recipient/sender pair, not just the
1968 recipient address. There doesn't seem to be a real danger of callout loops,
1969 since a callout by the remote host to check the sender would use <>.
1970 [SEE ABOVE: changed after hitting problems.]
1972 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1973 temporary errors. However, in the case of such a code being given after
1974 the end of a data transmission (i.e. after ".") Exim was failing to write
1975 a retry record for the message. (Yes, there was some broken host that was
1976 actually sending 8xx at this point.)
1978 73. An unknown lookup type in a host list could cause Exim to panic-die when
1979 the list was checked. (An example that provoked this was putting <; in the
1980 middle of a list instead of at the start.) If this happened during a DATA
1981 ACL check, a -D file could be left lying around. This kind of configuration
1982 error no longer causes Exim to die; instead it causes a defer errror. The
1983 incident is still logged to the main and panic logs.
1985 74. Buglet left over from Exim 3 conversion. The message "too many messages
1986 in one connection" was written to the rejectlog but not the mainlog, except
1987 when address rewriting (yes!) was being logged.
1989 75. Added write_rejectlog option.
1991 76. When a system filter was run not as root (that is, when system_filter_user
1992 was set), the values of the $n variables were not being returned to the
1993 main process; thus, they were not subsequently available in the $sn
1996 77. Added +return_path_on_delivery log selector.
1998 78. A connection timeout was being treated differently from recipients deferred
1999 when testing hosts_max_try with a message that was older than the host's
2000 retry timeout. (The host should not be counted, thus allowing all hosts to
2001 be tried at least once before bouncing.) This may have been the cause of an
2002 occasionally reported bug whereby a message would remain on the queue
2003 longer than the retry timeout, but would be bounced if a delivery was
2004 forced. I say "may" because I never totally pinned down the problem;
2005 setting up timeout/retry tests is difficult. See also the next item.
2007 79. The ultimate address timeout was not being applied to errors that involved
2008 a combination of host plus message (for example, a timeout on a MAIL
2009 command). When an address resolved to a number of possible hosts, and they
2010 were not all tried for each delivery (e.g. because of hosts_max_try), a
2011 message could remain on the queue longer than the retry timeout.
2013 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2016 81. Fixed an obscure SMTP outgoing bug which required at least the following
2017 conditions: (a) there was another message waiting for the same server;
2018 (b) the server returned 5xx to all RCPT commands in the first message so
2019 that the message was not completed; (c) the server dropped the connection
2020 or gave a negative response to the RSET that Exim sends to abort the
2021 transaction. The observed case was a dropped connection after DATA that had
2022 been sent in pipelining mode. That is, the server had advertised PIPELINING
2023 but was not implementing it correctly. The effect of the bug was incorrect
2024 behaviour, such as trying another host, and this could lead to a crash.
2030 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2031 and daemon.c were passed as pointers to ints; they should have been
2032 pointers to socklen_t variables (which are typically unsigned ints).
2034 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2037 3. Fixed a really odd bug that affected only the testing scheme; patching a
2038 certain fixed string in the binary changed the value of another string that
2039 happened to be identical to the end of the original first string.
2041 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2042 name", it returns that address as the IP address. On some operating
2043 systems (e.g. Solaris), it also passes back the IP address string as the
2044 "host name". However, on others (e.g. Linux), it passes back an empty
2045 string. Exim wasn't checking for this, and was changing the host name to an
2046 empty string, assuming it had been canonicized.
2048 5. Although rare, it is permitted to have more than one PTR record for a given
2049 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2050 all the names associated with an address, because they do in Solaris.
2051 However, it seems that they do not in Linux for data that comes from the
2052 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2053 I found this out when I moved to a new Linux workstation and tried to run
2054 the Exim test suite.
2056 To get round this problem I have changed the code so that it now does its
2057 own call to the DNS to look up PTR records when searching for a host name.
2058 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2059 addresses that are only in /etc/hosts are still found.
2061 This behaviour is, however, controlled by an option called host_lookup_
2062 order, which defaults to "bydns:byaddr". If people want to use the other
2063 order, or indeed, just use one or the other means of lookup, they can
2064 specify it in this variable.
2066 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2067 some operating systems, this comes back from gethostbyaddr() as an empty
2068 string, and this is what Exim used to test for. However, it seems that in
2069 other systems, "." is yielded. Exim now tests for this case too.
2071 7. The values of check_spool_space and check_log_space are now held internally
2072 as a number of kilobytes instead of an absolute number of bytes. If a
2073 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2074 kilobyte. This means that much larger values can be stored.
2076 8. Exim monitor: an attempt to get the action menu when not actually pointing
2077 at a message produces an empty menu entitled "No message selected". This
2078 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2079 no entries in it ("Shell widget menu has zero width and/or height"). So I
2080 have added a single, blank menu entry in this case.
2082 9. Added ${quote_local_part.
2084 10. MIME decoding is now applied to the contents of Subject: header lines when
2087 11. Now that a reference to $sender_host_address automatically causes a reverse
2088 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2089 host lookup before query-style lookups in lists that might use this
2090 variable. This has therefore been abolished, and the "net-" prefix is no
2091 longer necessary for query-style lookups.
2093 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2094 have been a typo for LFLAGS, so it has been changed.
2096 13. The install script calls Exim with "-C /dev/null" in order to find the
2097 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2098 to be output. Howeve, since Exim outputs its version number before the
2099 error, it didn't break the script. It just looked ugly. I fixed this by
2100 always allowing "-C /dev/null" if the caller is root.
2102 14. Ignore overlarge ACL variable number when reading spool file - insurance
2103 against a later release with more variables having written the file.
2105 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2106 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2107 this, as well as the form without the "IPv6" on the front.
2109 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2110 OS/Makefile-Darwin file.
2112 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2113 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2115 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2116 variable. However, some OS use ints for inodes. I've added cast to long int
2117 to get rid of the compiler warning.
2119 19. I had forgotten to lock out "/../" in configuration file names when
2120 ALT_CONFIG_PREFIX was set.
2122 20. Routers used for verification do not need to specify transports. However,
2123 if such a router generated a host list, and callout was configured, Exim
2124 crashed, because it could not find a port number from the (non-existent)
2125 transport. It now assumes port 25 in this circumstance.
2127 21. Added the -t option to exigrep.
2129 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2130 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2131 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2132 them is set, LOOKUP_LSEARCH is forced.
2134 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2136 24. Added sender and host information to the "rejected by local_scan()" log
2137 line; previously there was no indication of these.
2139 25. Added .include_if_exists.
2141 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2142 Linux. It turns out that not all file systems support this. Apparently some
2143 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2144 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2145 not supported on the file descriptor, is now ignored when Exim is trying to
2146 sync a directory. This applies only to Linux.
2148 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2150 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2151 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2154 29. Added support for maildirsize files from supplied patch (modified a bit).
2156 30. The use of :fail: followed by an empty string could lead Exim to respond to
2157 sender verification failures with (e.g.):
2159 550 Verification failed for <xxx>
2160 550 Sender verify failed
2162 where the first response line was missing the '-' that indicates it is not
2163 the final line of the response.
2165 31. The loop for finding the name of the user that called Exim had a hardwired
2166 limit of 10; it now uses the value of finduser_retries, which is used for
2167 all other user lookups.
2169 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2172 33. Exim was neglecting to zero errno before one call of strtol() when
2173 expanding a string and expecting an integer value. On some systems this
2174 resulted in spurious "integer overflow" errors. Also, it was casting the
2175 result into an int without checking.
2177 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2178 did not work. The code looks as if it has *never* worked, though it appears
2179 to have been documented since at least releast 1.62. I have made it work.
2181 35. The "timeout_DNS" error in retry rules, also documented since at least
2182 1.62, also never worked. As it isn't clear exactly what this means, and
2183 clearly it isn't a major issue, I have abolished the feature by treating it
2184 as "timeout", and writing a warning to the main and panic logs.
2186 36. The display of retry rules for -brt wasn't always showing the error code
2189 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2190 timeout_connect_A, timeout_connect_MX.
2192 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2193 to the empty sender.
2195 39. The daemon was not analysing the content of -oX till after it had closed
2196 stderr and disconnected from the controlling terminal. This meant that any
2197 syntax errors were only noted on the panic log, and the return code from
2198 the command was 0. By re-arranging the code a little, I've made the
2199 decoding happen first, so such errors now appear on stderr, and the return
2200 code is 1. However, the actual setting up of the sockets still happens in
2201 the disconnected process, so errors there are still only recorded on the
2204 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2205 connections (as happens on some IP stacks) was logged at start up time as
2206 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2207 it from "IPv6 and IPv4", which means that two separate sockets are being
2210 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2211 says whether AF_INET or AF_INET6 was passed as an argument.
2213 42. Exiwhat output was messed up when time zones were included in log
2216 43. Exiwhat now gives more information about the daemon's listening ports,
2217 and whether -tls-on-connect was used.
2219 44. The "port" option of the smtp transport is now expanded.
2221 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2222 silently ignored. Now an error message is written to the main and panic
2225 46. There's a new ACL modifier called "logwrite" which writes to a log file
2226 as soon as it is encountered.
2228 47. Added $local_user_uid and $local_user_gid at routing time.
2230 48. Exim crashed when trying to verify a sender address that was being
2233 49. Exim was recognizing only a space character after ".include". It now also
2234 recognizes a tab character.
2236 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2237 extracting the relevant information from the specification. The man page no
2238 longer contains scrambled data for the -d option, and I've added a section
2239 at the front about calling Exim under different names.
2241 51. Added "extra_headers" argument to the "mail" command in filter files.
2243 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2246 53. Installed eximstats 1.29.
2248 54. Added transport_filter_timeout as a generic transport option.
2250 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2251 Cc: header lines. This was required by RFC 822, but it not required by RFC
2254 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2255 incoming messages that did not have them. Now it does so only if the
2256 message originates locally, that is, if there is no associated remote host
2257 address. When Resent- header lines are present, this applies to the Resent-
2258 lines rather than the non-Resent- lines.
2260 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2261 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2263 58. Messages for configuration errors now include the name of the main
2264 configuration file - useful now that there may be more than one file in a
2265 list (.included file names were always shown).
2267 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2268 for those rare installations that do not start the daemon as root or run it
2269 setuid root. I've cut out the call to initgroups() if the daemon is not
2272 60. The Exim user and group can now be bound into the binary as text strings
2273 that are looked up at the start of Exim's processing.
2275 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2277 62. Added $mailstore_basename variable.
2279 63. Installed patch to sieve.c from Michael Haardt.
2281 64. When Exim failed to open the panic log after failing to open the main log,
2282 the original message it was trying to log was written to stderr and debug
2283 output, but if they were not available (the usual case in production), it
2284 was lost. Now it is written to syslog before the two lines that record the
2285 failures to open the logs.
2287 65. Users' Exim filters run in subprocesses under the user's uid. It is
2288 possible for a "deliver" command or an alias in a "personal" command to
2289 provoke an address rewrite. If logging of address rewriting is configured,
2290 this fails because the process is not running as root or exim. There may be
2291 a better way of dealing with this, but for the moment (because 4.30 needs
2292 to be released), I have disabled address rewrite logging when running a
2293 filter in a non-root, non-exim process.
2299 1. The buildconfig auxiliary program wasn't quoting the value set for
2300 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2301 not defined. This bug was masked in 4.22 by the effect that was fixed in
2304 2. Some messages that were rejected after a message id was allocated were
2305 shown as "incomplete" by exigrep. It no longer does this for messages that
2306 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2308 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2309 have allow_domain_literals set, the ID did not get logged in the <= line.
2310 Domain literals are now always recognized in Message-ID: header lines.
2312 4. The first argument for a ${extract expansion item is the key name or field
2313 number. Leading and trailing spaces in this item were not being ignored,
2314 causing some misleading effects.
2316 5. When deliver_drop_privilege was set, single queue runner processes started
2317 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2318 same command in the process it spins off) were not dropping privilege.
2320 6. When the daemon running as "exim" started a queue runner, it always
2321 re-executed Exim in the spun-off process. This is a waste of effort when
2322 deliver_drop_privilege is set. The new process now just calls the
2323 queue-runner function directly.
2329 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2332 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2333 ignored. Though the use of -oP was forcing the writing of a pid file, it
2334 was always written to the default place.
2336 3. If the message "no IP address found for host xxxx" is generated during
2337 incoming verification, it is now followed by identification of the incoming
2338 connection (so you can more easily find what provoked it).
2340 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2342 5. Added some features to "harden" Exim a bit more against certain attacks:
2344 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2345 be put in Local/Makefile. This is like the never_users runtime option,
2346 but it cannot be overridden. The default setting is "root".
2348 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2349 prefix string with which any file named in a -C command line option
2352 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2353 is retained for -C and -D only if the caller of Exim is root. Without
2354 it, the exim user may also use -C and -D and retain privilege.
2356 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2357 command line option is disabled.
2359 6. Macro names set by the -D option must start with an upper case letter, just
2360 like macro names defined in the configuration file.
2362 7. Added "dereference=" facility to LDAP.
2364 8. Two instances of the typo "uknown" in the source files are fixed.
2366 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2367 the Configure-Makefile script screwed up while processing it.
2369 10. Incorporated PCRE 4.4.
2371 11. The SMTP synchronization check was not operating right at the start of an
2372 SMTP session. For example, it could not catch a HELO sent before the client
2373 waited for the greeting. There is now a check for outstanding input at the
2374 point when the greeting is written. Because of the duplex, asynchronous
2375 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2376 way, but not yet received, when the check is performed.
2378 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2379 on TCP/IP sockets, because this apparently causes some broken clients to
2382 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2383 unchanged) from the Cygwin maintainer.
2385 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2386 was supported instead of testing for mbx. Effectively a typo.
2388 15. The spa authenticator server code was not checking that the input it
2389 received was valid base64.
2391 16. The debug output line for the "set" modifier in ACLs was not showing the
2392 name of the variable that was being set.
2394 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2396 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2397 lookup on its own. Something else was needed to trigger the lookup. For
2398 example, a match in host_lookup or the need for a host name in a host list.
2399 Now, if $sender_host_name is referenced and the host name has not yet been
2400 looked up, a lookup is performed. If the lookup fails, the variable remains
2401 empty, and $host_lookup_failed is set to "1".
2403 19. Added "eqi" as a case-independent comparison operator.
2405 20. The saslauthd authentication condition could segfault if neither service
2406 nor realm was specified.
2408 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2409 error message that was logged was misleading, and incoming SMTP
2410 connections were dropped. The message is now more accurate, and temporary
2411 errors are given to SMTP connections.
2413 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2414 (except RSET) with a 421 error, until QUIT is received. However, it was
2415 failing to send a response to QUIT.
2417 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2418 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2419 ACL is now run after the helo_try_verify_hosts code.
2421 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2422 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2423 case-independent, so other case variants are also recognized). Apparently
2424 some systems use these upper case variants.
2426 25. If more than two messages were waiting for the same host, and a transport
2427 filter was specified for the transport, Exim sent two messages over the
2428 same TCP/IP connection, and then failed with "socket operation on non-
2429 socket" when it tried to send the third.
2431 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2433 27. The extern definition of crypt16() in expand.c was not being excluded when
2434 the OS had its own crypt16() function.
2436 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2437 as a preferred synonym for return_size_limit, both as an option and as an
2440 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2442 30. Changed the default configuration ACL to relax the local part checking rule
2443 for addresses that are not in any local domains. For these addresses,
2444 slashes and pipe symbols are allowed within local parts, but the sequence
2445 /../ is explicitly forbidden.
2447 31. SPA server authentication was not clearing the challenge buffer before
2450 32. log_message in a "warn" ACL statement was writing to the reject log as
2451 well as to the main log, which contradicts the documentation and doesn't
2452 seem right (because no rejection is happening). So I have stopped it.
2454 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2455 However, I am unable to do any testing of this.
2457 34. Fixed an infelicity in the appendfile transport. When checking directories
2458 for a mailbox, to see if any needed to be created, it was accidentally
2459 using path names with one or more superfluous leading slashes; tracing
2460 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2462 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2463 contents are added to the log line that is written for every discarded
2464 recipient. (Previously a log_message setting was ignored.)
2466 36. The ${quote: operator now quotes the string if it is empty.
2468 37. The install script runs exim in order to find its version number. If for
2469 some reason other than non-existence or emptiness, which it checks, it
2470 could not run './exim', it was installing it with an empty version number,
2471 i.e. as "exim-". This error state is now caught, and the installation is
2474 38. An argument was missing from the function that creates an error message
2475 when Exim fails to connect to the socket for saslauthd authentication.
2476 This could cause Exim to crash, or give a corrupted message.
2478 39. Added isip, isip4, and isip6 to ${if conditions.
2480 40. The ACL variables $acl_xx are now saved with the message, and can be
2481 accessed later in routers, transports, and filters.
2483 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2484 strings in the file are not string-expanded.
2486 42. If a MAIL command specified a SIZE value that was too large to fit into an
2487 int variable, the check against message_size_limit failed. Such values are
2488 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2489 day this will have to be increased, but I don't think I want to be around
2490 when emails are that large.
2497 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2498 iconv() is not standard in FreeBSD.
2500 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2501 IPv6 enabled. The observed symptom was a segmentation fault on return from
2502 the function os_common_find_running_interfaces() in src/os.c.
2504 3. In the check_special_case() function in daemon.c I had used "errno" as an
2505 argument name, which causes warnings on some systems. This was basically a
2506 typo, since it was named "eno" in the comments!
2508 4. The code that waits for the clock to tick (at a resolution of some fraction
2509 of a second) so as to ensure message-id uniqueness was always waiting for
2510 at least one whole tick, when it could have waited for less. [This is
2511 almost certainly not relevant at current processor speeds, where it is
2512 unlikely to ever wait at all. But we try to future-proof.]
2514 5. The function that sleeps for a time interval that includes fractions of a
2515 second contained a race. It did not block SIGALRM between setting the
2516 timer, and suspending (a couple of lines later). If the interval was short
2517 and the sigsuspend() was delayed until after it had expired, the suspension
2518 never ended. On busy systems this could lead to processes getting stuck for
2521 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2522 process, before it forks any delivery processes. The open lookup caching
2523 mechanism meant that the open file or database connection was passed into
2524 the delivery process. The problem was that delivery processes always tidy
2525 up cached lookup data. This could cause a problem for the next delivery
2526 process started by the queue runner, because the external queue runner
2527 process does not know about the closure. So the next delivery process
2528 still has data in the lookup cache. In the case of a file lookup, there was
2529 no problem because closing a file descriptor in a subprocess doesn't affect
2530 the parent. However, if the lookup was caching a connection to a database,
2531 the connection was closed, and the second delivery process was likely to
2532 see errors such as "PGSQL: query failed: server closed the connection
2533 unexpectedly". The problem has been fixed by closing all cached lookups
2534 in a queue runner before running a delivery process.
2536 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2537 seem to have the "const" qualifier which it has on other OS. I've
2540 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2541 *of the same type* (client or server) with the same public name that an
2542 error should be diagnosed.
2544 9. When Exim looked up a host name for an IP address, but failed to find the
2545 original IP address when looking up the host name (a safety check), it
2546 output the message "<ip address> does not match any IP for NULL", which was
2547 confusing, to say the least. The bug was that the host name should have
2548 appeared instead of "NULL".
2550 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2551 user that is built into the binary, and the -C or -D options is used, root
2552 privilege is dropped before the configuration file is read. In addition,
2553 logging is switched to stderr instead of the normal log files. If the
2554 configuration then re-defines the Exim user, the unprivileged environment
2555 is probably not what is expected, so Exim logs a panic warning message (but
2558 However, if deliver_drop_privilege is set, the unprivileged state may well
2559 be exactly what is intended, so the warning has been cut out in that case,
2560 and Exim is allowed to try to write to its normal log files.
2566 1. smtp_return_error_details was not giving details for temporary sender
2567 or receiver verification errors.
2569 2. Diagnose a configuration error if two authenticators have the same public
2572 3. Exim used not to create the message log file for a message until the first
2573 delivery attempt. This could be confusing when incoming messages were held
2574 for policy or load reasons. The message log file is now created at the time
2575 the message is received, and an initial "Received" line is written to it.
2577 4. The automatically generated man page for command line options had a minor
2578 bug that caused no ill effects; however, a more serious problem was that
2579 the procedure for building the man page automatically didn't always
2580 operate. Consequently, release 4.20 contains an out-of-date version. This
2581 shouldn't happen again.
2583 5. When building Exim with embedded Perl support, the script that builds the
2584 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2585 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2587 6. The freeze_tell option was not being used for messages that were frozen on
2588 arrival, either by an ACL or by local_scan().
2590 7. Added the smtp_incomplete_transaction log selector.
2592 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2593 was accepting AUTH without a new EHLO.
2595 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2596 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2599 10. Logging of TCP/IP connections (when configured) now happens in the main
2600 daemon process instead of the child process, so that the TCP/IP connection
2601 count is more accurate (but it can never be perfect).
2603 11. The use of "drop" in a nested ACL was not being handled correctly in the
2604 outer ACL. Now, if condition failure induced by the nested "drop" causes
2605 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2606 or "require"), the connection is dropped.
2608 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2609 that yield "discard" can now be used with an "accept" or a "discard" verb,
2610 but an error is generated for any others (because I can't see a useful way
2611 to define what should happen).
2613 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2614 are now processed in the same way as lines in the Exim configuration file.
2615 In particular, continuation lines are supported.
2617 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2619 15. Added -ti meaning -t -i.
2621 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2622 domains, and warn by logging if others are found.
2624 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2625 returned from GIFCONF when Exim is trying to find the list of interfaces on
2626 a host. The code in os.c has been modified to copy each ifreq to an aligned
2627 structure in all cases.
2629 Also, in some cases, the returned ifreq's were being copied to a 'struct
2630 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2631 means the last couple of bytes of an IPv6 address could be chopped if the
2632 ifreq contained only a normal sockaddr (14 bytes storage).
2634 18. Named domain lists were not supported in the hosts_treat_as_local option.
2635 An entry such as +xxxx was not recognized, and was treated as a literal
2638 19. Ensure that header lines added by a DATA ACL are included in the reject log
2639 if the ACL subsequently rejects the message.
2641 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2642 MD5 (which is deprecated).
2644 21. When testing a filter file using -bf, Exim was writing a message when it
2645 took the sender from a "From " line in the message, but it was not doing so
2646 when it took $return_path from a Return-Path: header line. It now does.
2648 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2649 with a valid header line field name (a series of printing characters
2650 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2652 23. Changed "disc" in the source to "disk" to conform to the documentation and
2653 the book and for uniformity.
2655 24. Ignore Sendmail's -Ooption=value command line item.
2657 25. When execve() failed while trying to run a command in a pipe transport,
2658 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2659 could be confused with a return value of 69 from the command itself. This
2660 has been changed to 127, the value the shell returns if it is asked to run
2661 a non-existent command. The wording for the related log line suggests a
2662 non-existent command as the problem.
2664 26. If received_header_text expands to an empty string, do not add a Received:
2665 header line to the message. (Well, it adds a token one on the spool, but
2666 marks it "old" so that it doesn't get used or transmitted.)
2668 27. Installed eximstats 1.28 (addition of -nt option).
2670 28. There was no check for failure on the call to getsockname() in the daemon
2671 code. This can fail if there is a shortage of resources on the system, with
2672 ENOMEM, for example. A temporary error is now given on failure.
2674 29. Contrary to the C standard, it seems that in some environments, the
2675 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2676 program. Exim now does this explicitly; it affects the formatting of
2677 timestamps using strftime().
2679 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2680 complaints. I've now initialized all the variables, to avoid this.
2682 32. Header lines added by a system filter were not being "seen" during
2683 transport-time rewrites.
2685 33. The info_callback() function passed to OpenSSL is set up with type void
2686 (*)(SSL *, int, int), as described somewhere. However, when calling the
2687 function (actually a macro) that sets it up, the type void(*)() is
2688 expected. I've put in a cast to prevent warnings from picky compilers.
2690 34. If a DNS black list lookup found a CNAME record, but there were no A
2691 records associated with the domain it pointed at, Exim crashed.
2693 35. If a DNS black list lookup returned more than one A record, Exim ignored
2694 all but the first. It now scans all returned addresses if a particular IP
2695 value is being sought. In this situation, the contents of the
2696 $dnslist_value variable are a list of all the addresses, separated by a
2699 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2700 to accept a host name and all its aliases if the forward lookup for any of
2701 them yielded the IP address of the incoming connection. Now it accepts only
2702 those names whose forward lookup yields the correct IP address. Any other
2703 names are discarded. This closes a loophole whereby a rogue DNS
2704 administrator could create reverse DNS records to break through a
2705 wildcarded host restriction in an ACL.
2707 37. If a user filter or a system filter that ran in a subprocess used any of
2708 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2709 the wrong values were passed to the pipe command ($thisaddress had the
2710 value of $0, $0 had the value of $1, etc). This bug was introduced by
2711 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2713 38. Improved the line breaking for long SMTP error messages from ACLs.
2714 Previously, if there was no break point between 40 and 75 characters, Exim
2715 left the rest of the message alone. Two changes have been made: (a) I've
2716 reduced the minimum length to 35 characters; (b) if it can't find a break
2717 point between 35 and 75 characters, it looks ahead and uses the first one
2718 that it finds. This may give the occasional overlong line, but at least the
2719 remaining text gets split now.
2721 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2722 file descriptors might be low, and that setting 1000 would always raise it.
2723 It turns out that in some environments, the limit is already over 1000 and
2724 that lowering it causes trouble. So now Exim takes care not to decrease it.
2726 40. When delivering a message, the value of $return_path is set to $sender_
2727 address at the start of routing (routers may change the value). By an
2728 oversight, this default was not being set up when an address was tested by
2729 -bt or -bv, which affected the outcome if any router or filter referred to
2732 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2733 log only when "message" or "log_message" are set. However, if one of the
2734 conditions was an address verification, or a call to a nested ACL, the
2735 messages generated by the underlying test were being passed through. This
2736 no longer happens. The underlying message is available in $acl_verify_
2737 message for both "message" and "log_message" expansions, so it can be
2738 passed through if needed.
2740 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2741 new expansion $bh_ to give the encoded byte string without charset
2742 translation. Translation happens only if iconv() is available; HAVE_ICONV
2743 indicates this at build time. HEADERS_CHARSET gives the charset to
2744 translate to; headers_charset can change it in the configuration, and
2745 "headers charset" can change it in an individual filter file.
2747 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2748 that creates RFC 2047 encoded "words" labels them as that charset instead
2749 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2750 expansion operator; (ii) when Exim creates a From: line for a local
2751 message; (iii) when a header line is rewritten to include a "phrase" part.
2753 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2754 buggy, causing it to skip the first lines of messages whose message ID
2755 ended in 'D'. This would not have bitten before Exim release 4.14, because
2756 message IDs were unlikely to end in 'D' before then. The effect was to have
2757 incorrect size information for certain domains.
2759 45. #include "config.h" was missing at the start of the crypt16.c module. This
2760 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2763 46. If there was a timeout during a "random" callout check, Exim treated it as
2764 a failure of the random address, and carried on sending RSET and the real
2765 address. If the delay was just some slowness somewhere, the response to the
2766 original RCPT would be taken as a response to RSET and so on, causing
2767 mayhem of various kinds.
2769 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2770 when I implemented it. It didn't allow for the fact that some option values
2771 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2772 the right test for positive values.
2774 48. Domain names in DNS records are case-independent. Exim always looks them up
2775 in lower case. Some resolvers return domain names in exactly the case they
2776 appear in the zone file, that is, they may contain uppercase letters. Not
2777 all resolvers do this - some return always lower case. Exim was treating a
2778 change of case by a resolver as a change of domain, similar to a widening
2779 of a domain abbreviation. This triggered its re-routing code and so it was
2780 trying to route what was effectively the same domain again. This normally
2781 caused routing to fail (because the router wouldn't handle the domain
2782 twice). Now Exim checks for this case specially, and just changes the
2783 casing of the domain that it ultimately uses when it transmits the message
2786 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2789 50. If a filter generated a file delivery with a non-absolute name (possible if
2790 no home directory exists for the router), the forbid_file option was not
2793 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2794 the existing equality matching.
2796 52. Exim was using ints instead of ino_t variables in some places where it was
2797 dealing with inode numbers.
2799 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2800 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2801 called TMPDIR, and if it finds it is different, it changes its value.
2803 54. The smtp_printf() function is now made available to local_scan() so
2804 additional output lines can be written before returning. There is also an
2805 smtp_fflush() function to enable the detection of a dropped connection.
2806 The variables smtp_input and smtp_batched_input are exported to
2809 55. Changed the default runtime configuration: the message "Unknown user"
2810 has been removed from the ACL, and instead placed on the localuser router,
2811 using the cannot_route_message feature. This means that any verification
2812 failures that generate their own messages won't get overridden. Similarly,
2813 the "Unrouteable address" message that was in the ACL for unverifiable
2814 relay addresses has also been removed.
2816 56. Added hosts_avoid_esmtp to the smtp transport.
2818 57. The exicyclog script was not checking for the esoteric option
2819 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2820 will work only if exicyclog is run under the appropriate euid.
2822 58. Following a discussion on the list, the rules by which Exim recognises line
2823 endings on incoming messages have been changed. The -dropcr and drop_cr
2824 options are now no-ops, retained only for backwards compatibility. The
2825 following line terminators are recognized: LF CRLF CR. However, special
2826 processing applies to CR:
2828 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2829 nor a local message in the state where . is a terminator.
2831 (ii) If a bare CR is encountered in a header line, an extra space is added
2832 after the line terminator so as not to end the header. The reasoning
2833 behind this is that bare CRs in header lines are most likely either
2834 to be mistakes, or people trying to play silly games.
2836 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2837 was being incorrectly given as 18 bytes larger than it should have been.
2838 This is a VOB (very old bug).
2840 60. This may never have affected anything current, but just in case it has:
2841 When the local host is found other than at the start of a list of hosts,
2842 the local host, those with the same MX, and any that follow, are discarded.
2843 When the list in question was part of a longer list of hosts, the following
2844 hosts (not currently being processed) were also being discarded. This no
2845 longer happens. I'm not sure if this situation could ever has previously
2848 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2851 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2852 header. This is something that is recommended in a new Internet Draft, and
2853 is something that is documented as being done by Sendmail. There are two
2854 possible values. For messages generated by the autoreply transport, Exim
2857 Auto-Submitted: auto-replied
2859 whereas for all other generated messages (e.g. bounces) it adds
2861 Auto-Submitted: auto-generated
2863 63. The "personal" condition in filters now includes a test for the
2864 Auto-Submitted: header. If it contains the string "auto-" the message it
2865 not considered personal.
2867 64. Added rcpt_include_affixes as a generic transport option.
2869 65. Added queue_only_override (default true).
2871 66. Added the syslog_duplication option.
2873 67. If what should have been the first header line of a message consisted of
2874 a space followed by a colon, Exim was mis-interpreting it as a header line.
2875 It isn't of course - it is syntactically invalid and should therefore be
2876 treated as the start of the message body. The misbehaviour could have
2877 caused a number of strange effects, including loss of data in subsequent
2878 header lines, and spool format errors.
2880 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2881 client host had authenticated. This control can now be exercised by an ACL
2882 for more flexibility.
2884 69. By default, callouts do not happen when testing with -bh. There is now a
2885 variant, -bhc, which does actually run the callout code, including
2886 consulting and updating the callout cache.
2888 70. Added support for saslauthd authentication, courtesy of Alexander
2891 71. If statvfs() failed on the spool or log directories while checking their
2892 size for availability, Exim confusingly gave the error "space shortage".
2893 Furthermore, in debugging mode it crashed with a floating point exception.
2894 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2895 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2896 serious problem, Exim now writes to the main and panic logs when this
2897 happens, with details of the failure. It then refuses to accept the
2898 incoming message, giving the message "spool directory problem" or "log
2899 directory problem" with a 421 code for SMTP messages.
2901 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2902 0, 1, and 2 exist, because some OS complain for execs without them (see
2903 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2904 descriptors. However, the code omitted to check that the open succeeded,
2905 causing mysterious errors if for some reason the permissions on /dev/null
2906 got screwed. Now Exim writes a message to the main and panic logs, and
2907 bombs out if it can't open /dev/null.
2909 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2910 interact so that it is all more flexible. It is supposed to remain
2911 backwards compatible. Also added extra_local_interfaces.
2913 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2914 to bomb out with an assertion failure - to the client this appears as a
2915 connection drop. This problem occurs in the part of the code that was taken
2916 from the Samba project. Fortunately, the assertion is in a very simple
2917 function, so I have fixed this by reproducing the function inline in the
2918 one place where it is called, and arranging for authentication to fail
2919 instead of killing the process with assert().
2921 75. The SPA client code was not working when the server requested OEM rather
2922 than Unicode encoding.
2924 76. Added code to make require_files with a specific uid setting more usable in
2925 the case where statting the file as root fails - usually a non-root-mounted
2926 NFS file system. When this happens and the failure is EACCES, Exim now
2927 forks a subprocess and does the per-uid checking as the relevant uid.
2929 77. Added process_log_path.
2931 78. If log_file_path was not explicitly set, a setting of check_log_space or
2932 check_log_inodes was ignored.
2934 79. If a space check for the spool or log partitions fails, the incident is now
2935 logged. Of course, in the latter case the data may get lost...
2937 80. Added the %p formatting code to string_format() so that it can be used to
2938 print addresses in debug_print(). Adjusted all the address printing in the
2939 debugging in store.c to use %p rather than %d.
2941 81. There was a concern that a line of code in smtp_in.c could overflow a
2942 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2943 initially expressed, the concern was not well-founded, because trailing
2944 spaces are removed early. However, if the trailing spaces were followed by
2945 a NULL, they did not get removed, so the overflow was possible. Two fixes
2948 (a) I re-wrote the offending code in a cleaner fashion.
2949 (b) If an incoming SMTP command contains a NULL character, it is rejected
2952 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2953 runs initgroups(), so that if the Exim user is in any additional groups,
2954 they will be used during message reception.
2960 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob