1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.146 2005/05/31 10:58:18 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.
51 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
52 in the smtp transport. When a block could not be written in a single
53 write() function, the timeout was being re-applied to each part-write.
54 This seems wrong - if the receiver was accepting one byte at a time it
55 would take for ever. The timeout is now adjusted when this happens. It
56 doesn't have to be particularly precise.
58 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
59 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
61 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
62 without a domain if the check to <postmaster@domain> fails.
68 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
69 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
71 TK/02 Fix ACL "control" statement not being available in MIME ACL.
73 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
75 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
76 to test Sieve filters that use "vacation".
78 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
79 that changes the way the GnuTLS parameters are stored in the cache file.
80 The new format can be generated externally. For backward compatibility,
81 if the data in the cache doesn't make sense, Exim assumes it has read an
82 old-format file, and it generates new data and writes a new file. This
83 means that you can't go back to an older release without removing the
86 PH/03 A redirect router that has both "unseen" and "one_time" set does not
87 work if there are any delivery delays because "one_time" forces the
88 parent to be marked "delivered", so its unseen clone is never tried
89 again. For this reason, Exim now forbids the simultaneous setting of
92 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
93 redirected to themselves ("homonym" addresses). Read the long ChangeLog
94 entry if you want to know the details. The fix, however, neglected to
95 consider the case when local delivery batching is involved. The test for
96 "previously delivered" was not happening when checking to see if an
97 address could be batched with a previous (undelivered) one; under
98 certain circumstances this could lead to multiple deliveries to the same
101 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
102 in its include files, and this causes problems building Exim.
104 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
105 header_syntax) but Exim was just ignoring anything given after a slash.
106 In particular, this caused confusion with an attempt to use "verify =
107 reverse_host_lookup/defer_ok". An error is now given when options are
108 supplied for verify items that do not have them. (Maybe reverse_host_
109 lookup should have a defer_ok option, but that's a different point.)
111 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
112 defined by RFC 821) to 2048, because there were problems with some AUTH
113 commands, and RFC 1869 says the size should be increased for extended
114 SMTP commands that take arguments.
116 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
119 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
120 "unknown" error; now it says that the functionality isn't in the binary.
122 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
123 an address' error message when a string expansion fails (syntax or
124 whatever). Otherwise the password may appear in the log. Following change
125 PH/42 below, there is no longer a chance of it appearing in a bounce
128 PH/11 Installed exipick version 20050225.0 from John Jetmore.
130 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
131 of its addresses was ever tried. (Bugzilla bug #2.)
133 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
134 the result incorrectly in the debug output. (It correctly added a newline
135 to what was transported.)
137 TF/01 Added $received_time.
139 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
140 commented out examples of how to interface to a virus scanner and to
141 SpamAssassin. Also added commented examples of av_scanner and
142 spamd_address settings.
144 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
145 and controls are allowed in which ACLs. There were a couple of minor
146 errors. Some of the entries in the conditions table (which is a table of
147 where they are NOT allowed) were getting very unwieldy; rewrote them as a
148 negation of where the condition IS allowed.
150 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
152 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
153 header file does not have a version number, so I've had to invent a new
154 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
155 API. The code is untested by me (my Linux distribution still has 0.3.2 of
156 radiusclient), but it was contributed by a Radius user.
158 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
159 files or directories, for OpenSSL.
161 PH/19 When an Exim process that is running as root has to create an Exim log
162 file, it does so in a subprocess that runs as exim:exim so as to get the
163 ownership right at creation (otherwise, other Exim processes might see
164 the file with the wrong ownership). There was no test for failure of this
165 fork() call, which would lead to the process getting stuck as it waited
166 for a non-existent subprocess. Forks do occasionally fail when resources
167 run out. I reviewed all the other calls to fork(); they all seem to check
170 PH/20 When checking for unexpected SMTP input at connect time (before writing
171 the banner), Exim was not dealing correctly with a non-positive return
172 from the read() function. If the client had disconnected by this time,
173 the result was a log entry for a synchronization error with an empty
174 string after "input=" when read() returned zero. If read() returned -1
175 (an event I could not check), uninitialized data bytes were printed.
176 There were reports of junk text (parts of files, etc) appearing after
179 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
181 PH/22 Added support for macro redefinition, and (re)definition in between
182 driver and ACL definitions.
184 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
185 forgetting to use the resulting value; it was using the unexpanded value.
187 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
188 hadn't been configured. The fix is from Juergen Kreileder, who
189 understands it better than I do:
191 "Here's what I see happening with three configured cyrus_sasl
192 authenticators configured (plain, login, cram-md5):
194 On startup auth_cyrus_sasl_init() gets called for each of these.
195 This means three calls to sasl_listmech() without a specified mech_list.
196 => SASL tests which mechs of all available mechs actually work
197 => three warnings about OTP not working
198 => the returned list contains: plain, login, cram-md5, digest-md5, ...
200 With the patch, sasl_listmech() also gets called three times. But now
201 SASL's mech_list option is set to the server_mech specified in the the
202 authenticator. Or in other words, the answer from sasl_listmech()
203 gets limited to just the mech you're testing for (which is different
205 => the return list contains just 'plain' or 'login', 'cram-md5' or
206 nothing depending on the value of ob->server_mech.
208 I've just tested the patch: Authentication still works fine,
209 unavailable mechs specified in the exim configuration are still
210 caught, and the auth.log warnings about OTP are gone."
212 PH/25 When debugging is enabled, the contents of the command line are added
213 to the debugging output, even when log_selector=+arguments is not
216 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
217 answer is "GNU", and only if the return is "GNU/something" is the answer
220 PH/27 $acl_verify_message is now set immediately after the failure of a
221 verification in an ACL, and so is available in subsequent modifiers. In
222 particular, the message can be preserved by coding like this:
224 warn !verify = sender
225 set acl_m0 = $acl_verify_message
227 Previously, $acl_verify_message was set only while expanding "message"
228 and "log_message" when a very denied access.
230 PH/28 Modified OS/os.c-Linux with
232 -#ifndef OS_LOAD_AVERAGE
233 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
235 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
236 nomenclature these days.)
238 PH/29 Installed patch from the Sieve maintainer that adds the options
239 sieve_useraddress and sieve_subaddress to the redirect router.
241 PH/30 In these circumstances:
242 . Two addresses routed to the same list of hosts;
243 . First host does not offer TLS;
244 . First host accepts first address;
245 . First host gives temporary error to second address;
246 . Second host offers TLS and a TLS session is established;
247 . Second host accepts second address.
248 Exim incorrectly logged both deliveries with the TLS parameters (cipher
249 and peerdn, if requested) that were in fact used only for the second
252 PH/31 When doing a callout as part of verifying an address, Exim was not paying
253 attention to any local part prefix or suffix that was matched by the
254 router that accepted the address. It now behaves in the same way as it
255 does for delivery: the affixes are removed from the local part unless
256 rcpt_include_affixes is set on the transport.
258 PH/32 Add the sender address, as F=<...>, to the log line when logging a
259 timeout during the DATA phase of an incoming message.
261 PH/33 Sieve envelope tests were broken for match types other than :is. I have
262 applied a patch sanctioned by the Sieve maintainer.
264 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
265 the uid or gid is negative. A case of a negative gid caused this to be
266 noticed. The fix allows for either to be negative.
268 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
269 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
272 PH/36 The change PH/12 above was broken. Fixed it.
274 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
275 the grounds that routing the same address twice would always produce the
276 same answer. This might have been true once, but it is certainly no
277 longer true now. Routing a child address may depend on the previous
278 routing that produced that child. Some complicated redirection strategies
279 went wrong when messages had multiple recipients, and made Exim's
280 behaviour dependent on the order in which the addresses were given.
282 I have moved the duplicate checking until after the routing is complete.
283 Exim scans the addresses that are assigned to local and remote
284 transports, and removes any duplicates. This means that more work will be
285 done, as duplicates will always all be routed, but duplicates are
286 presumably rare, so I don't expect this is of any significance.
288 For deliveries to pipes, files, and autoreplies, the duplicate checking
289 still happens during the routing process, since they are not going to be
292 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
293 It corrects a timeout issue with spamd. This is Ian's comment: "The
294 background is that sometimes spamd either never reads data from a
295 connection it has accepted, or it never writes response data. The exiscan
296 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
297 blindly assumes that writes won't block so it may never time out."
299 PH/39 Allow G after quota size as well as K and M.
301 PH/40 The value set for $authenticated_id in an authenticator may not contain
302 binary zeroes or newlines because the value is written to log lines and
303 to spool files. There was no check on this. Now the value is run through
304 the string_printing() function so that such characters are converted to
305 printable escape sequences.
307 PH/41 $message_linecount is a new variable that contains the total number of
308 lines in the message. Compare $body_linecount, which is the count for the
311 PH/42 Exim no longer gives details of delivery errors for specific addresses in
312 bounce and delay warning messages, except in certain special cases, which
315 (a) An SMTP error message from a remote host;
316 (b) A message specified in a :fail: redirection;
317 (c) A message specified in a "fail" command in a system filter;
318 (d) A message specified in a FAIL return from the queryprogram router;
319 (e) A message specified by the cannot_route_message router option.
321 In these cases only, Exim does include the error details in bounce and
322 warning messages. There are also a few cases where bland messages such
323 as "unrouteable address" or "local delivery error" are given.
325 PH/43 $value is now also set for the "else" part of a ${run expansion.
327 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
328 being worked on, but at least Exim now implements the latest version to
331 PH/45 In a pipe transport, although a timeout while waiting for the pipe
332 process to complete was treated as a delivery failure, a timeout while
333 writing the message to the pipe was logged, but erroneously treated as a
334 successful delivery. Such timeouts include transport filter timeouts. For
335 consistency with the overall process timeout, these timeouts are now
336 treated as errors, giving rise to delivery failures by default. However,
337 there is now a new Boolean option for the pipe transport called
338 timeout_defer, which, if set TRUE, converts the failures into defers for
339 both kinds of timeout. A transport filter timeout is now identified in
342 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
343 systems where "make" and "gmake" are different, calling "gmake" at top
344 level broke things. I've arranged for the value of $(MAKE) to be passed
345 from the Makefile to this script so that it can call the same version of
349 A note about Exim versions 4.44 and 4.50
350 ----------------------------------------
352 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
353 changes of various kinds. As a consequence, a big documentation update was
354 needed. This delayed the release for rather longer than seemed good, especially
355 in the light of a couple of (minor) security issues. Therefore, the changes
356 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
357 release. So 4.44 and 4.50 are in effect two different branches that both start
360 I have left the 4.50 change log unchanged; it contains all the changes since
361 4.43. The change log for 4.44 is below; many of its items are identical to
362 those for 4.50. This seems to be the most sensible way to preserve the
363 historical information.
369 1. Minor wording change to the doc/README.SIEVE file.
371 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
372 computation of the current number of files was incorrect.
374 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
375 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
378 4. Give more explanation in the error message when the command for a transport
379 filter fails to execute.
381 5. There are several places where Exim runs a non-Exim command in a
382 subprocess. The SIGUSR1 signal should be disabled for these processes. This
383 was being done only for the command run by the queryprogram router. It is
384 now done for all such subprocesses. The other cases are: ${run, transport
385 filters, and the commands run by the lmtp and pipe transports.
387 6. Added CONFIGURE_GROUP build-time option.
389 7. Some older OS have a limit of 256 on the maximum number of file
390 descriptors. Exim was using setrlimit() to set 1000 as a large value
391 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
392 systems. I've change it so that if it can't get 1000, it tries for 256.
394 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
395 was an oversight, and furthermore, ever since the addition of extra
396 controls (e.g. 4.43/32), the checks on when to allow different forms of
397 "control" were broken. There should now be diagnostics for all cases when a
398 control that does not make sense is encountered.
400 9. Added the /retain_sender option to "control=submission".
402 10. $recipients is now available in the predata ACL (oversight).
404 11. Tidy the search cache before the fork to do a delivery from a message
405 received from the command line. Otherwise the child will trigger a lookup
406 failure and thereby defer the delivery if it tries to use (for example) a
407 cached ldap connection that the parent has called unbind on.
409 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
410 of $address_data from the recipient verification was clobbered by the
413 13. The value of address_data from a sender verification is now available in
414 $sender_address_data in subsequent conditions in the ACL statement.
416 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
418 15. Added a new option "connect=<time>" to callout options, to set a different
421 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
422 was its contents. (It was OK if the option was not defined at all.)
424 17. A "Completed" log line is now written for messages that are removed from
425 the spool by the -Mrm option.
427 18. New variables $sender_verify_failure and $recipient_verify_failure contain
428 information about exactly what failed.
430 19. Added -dd to debug only the daemon process.
432 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
433 handles timeouts, both on the server side and network timeouts. Renamed the
434 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
436 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
437 It is now set to "smtps".
439 22. $host_address is now set to the target address during the checking of
442 23. When checking ignore_target_hosts for an ipliteral router, no host name was
443 being passed; this would have caused $sender_host_name to have been used if
444 matching the list had actually called for a host name (not very likely,
445 since this list is usually IP addresses). A host name is now passed as
448 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
449 code that specifies a non-restarting handler (typically sigaction() in
450 modern systems) in an attempt to fix a rare and obscure crash bug.
452 25. Narrowed the window for a race in the daemon that could cause it to ignore
453 SIGCHLD signals. This is not a major problem, because they are used only to
454 wake it up if nothing else does.
456 26. A malformed maildirsize file could cause Exim to calculate negative values
457 for the mailbox size or file count. Odd effects could occur as a result.
458 The maildirsize information is now recalculated if the size or filecount
461 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
462 support for a long time. Removed HAVE_SYS_VFS_H.
464 28. Installed the latest version of exipick from John Jetmore.
466 29. In an address list, if the pattern was not a regular expression, an empty
467 subject address (from a bounce message) matched only if the pattern was an
468 empty string. Non-empty patterns were not even tested. This was the wrong
469 because it is perfectly reasonable to use an empty address as part of a
470 database query. An empty address is now tested by patterns that are
471 lookups. However, all the other forms of pattern expect the subject to
472 contain a local part and a domain, and therefore, for them, an empty
473 address still always fails if the pattern is not itself empty.
475 30. Exim went into a mad DNS loop when attempting to do a callout where the
476 host was specified on an smtp transport, and looking it up yielded more
479 31. Re-factored the code for checking spool and log partition space into a
480 function that finds that data and another that does the check. The former
481 is then used to implement four new variables: $spool_space, $log_space,
482 $spool_inodes, and $log_inodes.
484 32. The RFC2047 encoding function was originally intended for short strings
485 such as real names; it was not keeping to the 75-character limit for
486 encoded words that the RFC imposes. It now respects the limit, and
487 generates multiple encoded words if necessary. To be on the safe side, I
488 have increased the buffer size for the ${rfc2047: expansion operator from
491 33. It is now permitted to omit both strings after an "if" condition; if the
492 condition is true, the result is "true". As before, when the second string
493 is omitted, a false condition yields an empty string. This makes it less
494 cumbersome to write custom ACL and router conditions.
496 34. Failure to deliver a bounce message always caused it to be frozen, even if
497 there was an errors_to setting on the router. The errors_to setting is now
500 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
501 canonical form (fully expanded) before being placed in
502 $sender_host_address.
504 36. The table in the code that translates DNS record types into text (T_A to
505 "A" for instance) was missing entries for NS and CNAME. It is just possible
506 that this could have caused confusion if both these types were looked up
507 for the same domain, because the text type is used as part of Exim's
508 per-process caching. But the chance of anyone hitting this buglet seems
511 37. The dnsdb lookup has been extended in a number of ways.
513 (1) There is a new type, "zns", which walks up the domain tree until it
514 finds some nameserver records. It should be used with care.
516 (2) There is a new type, "mxh", which is like "mx" except that it returns
517 just the host names, not the priorities.
519 (3) It is now possible to give a list of domains (or IP addresses) to be
520 looked up. The behaviour when one of the lookups defers can be
521 controlled by a keyword.
523 (4) It is now possible to specify the separator character for use when
524 multiple records are returned.
526 38. The dnslists ACL condition has been extended: it is now possible to supply
527 a list of IP addresses and/or domains to be looked up in a particular DNS
530 39. Added log_selector=+queue_time_overall.
532 40. When running the queue in the test harness, wait just a tad after forking a
533 delivery process, to get repeatability of debugging output.
535 41. Include certificate and key file names in error message when GnuTLS fails
536 to set them up, because the GnuTLS error message doesn't include the name
537 of the failing file when there is a problem reading it.
539 42. Allow both -bf and -bF in the same test run.
541 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
543 44. The "Exiscan patch" is now merged into the mainline Exim source.
545 45. Sometimes the final signoff response after QUIT could fail to get
546 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
547 < 0 before doing a fflush(). This bug looks as though it goes back to the
548 introduction of TLS in release 3.20, but "sometimes" must have been rare
549 because the tests only now provoked it.
551 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
552 (this can affect the format of dates).
554 47. exim_tidydb, when checking for the continued existence of a message for
555 which it has found a message-specific retry record, was not finding
556 messages that were in split spool directories. Consequently, it was
557 deleting retry records that should have stayed in existence.
559 48. Steve fixed some bugs in eximstats.
561 49. The SPA authentication driver was not abandoning authentication and moving
562 on to the next authenticator when an expansion was forced to fail,
563 contradicting the general specification for all authenticators. Instead it
564 was generating a temporary error. It now behaves as specified.
566 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
567 (the order specifies the preference for clients). The order is now AES256,
568 AES128, 3DES, ARCFOUR128.
570 51. Small patch to Sieve code - explicitly set From: when generating an
573 52. Exim crashed if a remote delivery caused a very long error message to be
574 recorded - for instance if somebody sent an entire SpamAssassin report back
575 as a large number of 550 error lines. This bug was coincidentally fixed by
576 increasing the size of one of Exim's internal buffers (big_buffer) that
577 happened as part of the Exiscan merge. However, to be on the safe side, I
578 have made the code more robust (and fixed the comments that describe what
581 53. Now that there can be additional text after "Completed" in log lines (if
582 the queue_time_overall log selector is set), a one-byte patch to exigrep
583 was needed to allow it to recognize "Completed" as not the last thing in
586 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
587 patch that reportedly fixes this has been added. I am not expert enough to
588 create a test for it. This is what the patch creator wrote:
590 "I found a little strange behaviour of ldap code when working with
591 Windows 2003 AD Domain, where users was placed in more than one
592 Organization Units. When I tried to give exim partial DN, the exit code
593 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
594 But simultaneously result of request was absolutely normal ldap result,
595 so I produce this patch..."
597 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
598 REFERENCE, so I have modified the code to exclude the patch when that macro
601 55. Some experimental protocols are using DNS PTR records for new purposes. The
602 keys for these records are domain names, not reversed IP addresses. The
603 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
604 leaves it alone. Component reversal etc. now happens only for IP addresses.
607 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
609 57. Double the size of the debug message buffer (to 2048) so that more of very
610 long debug lines gets shown.
612 58. The exicyclog utility now does better if the number of log files to keep
613 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
615 59. Two changes related to the smtp_active_hostname option:
617 (1) $smtp_active_hostname is now available as a variable.
618 (2) The default for smtp_banner uses $smtp_active_hostname instead
619 of $primary_hostname.
621 60. The host_aton() function is supposed to be passed a string that is known
622 to be a valid IP address. However, in the case of IPv6 addresses, it was
623 not checking this. This is a hostage to fortune. Exim now panics and dies
624 if the condition is not met. A case was found where this could be provoked
625 from a dnsdb PTR lookup with an IPv6 address that had more than 8
626 components; fortuitously, this particular loophole had already been fixed
627 by change 4.50/55 above.
629 If there are any other similar loopholes, the new check in host_aton()
630 itself should stop them being exploited. The report I received stated that
631 data on the command line could provoke the exploit when Exim was running as
632 exim, but did not say which command line option was involved. All I could
633 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
637 61. There was a buffer overflow vulnerability in the SPA authentication code
638 (which came originally from the Samba project). I have added a test to the
639 spa_base64_to_bits() function which I hope fixes it.
642 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
643 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
645 63. The daemon start-up calls getloadavg() while still root for those OS that
646 need the first call to be done as root, but it missed one case: when
647 deliver_queue_load_max is set with deliver_drop_privilege. This is
648 necessary for the benefit of the queue runner, because there is no re-exec
649 when deliver_drop_privilege is set.
651 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
654 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
655 was in use, was not putting the data itself into the right store pool;
656 consequently, it could be overwritten for a subsequent message in the same
657 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
660 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
662 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
663 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
664 the function were treating the return as a boolean value, which happened to
665 work because 0=false and not-0=true, but is not correct code.
667 68. The host_aton() function was not handling scoped IPv6 addresses (those
668 with, for example, "%eth0" on the end) correctly.
670 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
671 negated items (that is, ~something) in unsigned ints. Some compilers
672 apparently mutter when there is no cast.
674 70. If an address verification called from an ACL failed, and did not produce a
675 user-specific message (i.e. there was only a "system" message), nothing was
676 put in $acl_verify_message. In this situation, it now puts the system
679 71. Change 4.23/11 added synchronization checking at the start of an SMTP
680 session; change 4.31/43 added the unwanted input to the log line - except
681 that it did not do this in the start of session case. It now does.
683 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
684 This is wrong and can cause the other end to generate a synchronization
685 error if it is another Exim or anything else that does the synchronization
686 check. A QUIT command is no longer sent after a timeout.
688 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
691 74. The defer_ok option of callout verification was not working if it was used
692 when verifying addresses in header lines, that is, for this case:
694 verify = header_sender/callout=defer_ok
696 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
697 those file descriptors could be used for SMTP connections. If anything
698 wrote to stderr (the example that came up was "warn" in embedded Perl), it
699 could be sent to the SMTP client, causing chaos. The daemon now opens
700 stdin, stdout, and stderr to /dev/null when it puts itself into the
703 76. Arrange for output from Perl's "warn" command to be written to Exim's main
704 log by default. The user can override this with suitable Perl magic.
706 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
707 the log message when discard triggers, was not working for the DATA ACL or
708 for the non-SMTP ACL.
710 78. Error message wording change in sieve.c.
712 79. If smtp_accept_max_per_host was set, the number of connections could be
713 restricted to fewer than expected, because the daemon was trying to set up
714 a new connection before checking whether the processes handling previous
715 connections had finished. The check for completed processes is now done
716 earlier. On busy systems, this bug wouldn't be noticed because something
717 else would have woken the daemon, and it would have reaped the completed
720 80. If a message was submitted locally by a user whose login name contained one
721 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
722 It caused a spool format error. I have fixed the spool reading code. A
723 related problem was that the "from" clause in the Received: line became
724 illegal because of the space(s). It is now covered by ${quote_local_part.
726 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
729 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
731 83. Patch from Sieve maintainer to fix unterminated string problem in
734 84. Some minor changes to the Linux configuration files to help with other
735 OS variants using glibc.
737 85. One more patch for Sieve to update vacation handling to latest spec.
740 ----------------------------------------------------
741 See the note above about the 4.44 and 4.50 releases.
742 ----------------------------------------------------
748 1. Change 4.43/35 introduced a bug that caused file counts to be
749 incorrectly computed when quota_filecount was set in an appendfile
752 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
753 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
756 3. Give more explanation in the error message when the command for a transport
757 filter fails to execute.
759 4. There are several places where Exim runs a non-Exim command in a
760 subprocess. The SIGUSR1 signal should be disabled for these processes. This
761 was being done only for the command run by the queryprogram router. It is
762 now done for all such subprocesses. The other cases are: ${run, transport
763 filters, and the commands run by the lmtp and pipe transports.
765 5. Some older OS have a limit of 256 on the maximum number of file
766 descriptors. Exim was using setrlimit() to set 1000 as a large value
767 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
768 systems. I've change it so that if it can't get 1000, it tries for 256.
770 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
771 was an oversight, and furthermore, ever since the addition of extra
772 controls (e.g. 4.43/32), the checks on when to allow different forms of
773 "control" were broken. There should now be diagnostics for all cases when a
774 control that does not make sense is encountered.
776 7. $recipients is now available in the predata ACL (oversight).
778 8. Tidy the search cache before the fork to do a delivery from a message
779 received from the command line. Otherwise the child will trigger a lookup
780 failure and thereby defer the delivery if it tries to use (for example) a
781 cached ldap connection that the parent has called unbind on.
783 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
784 of $address_data from the recipient verification was clobbered by the
787 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
788 was its contents. (It was OK if the option was not defined at all.)
790 11. A "Completed" log line is now written for messages that are removed from
791 the spool by the -Mrm option.
793 12. $host_address is now set to the target address during the checking of
796 13. When checking ignore_target_hosts for an ipliteral router, no host name was
797 being passed; this would have caused $sender_host_name to have been used if
798 matching the list had actually called for a host name (not very likely,
799 since this list is usually IP addresses). A host name is now passed as
802 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
803 code that specifies a non-restarting handler (typically sigaction() in
804 modern systems) in an attempt to fix a rare and obscure crash bug.
806 15. Narrowed the window for a race in the daemon that could cause it to ignore
807 SIGCHLD signals. This is not a major problem, because they are used only to
808 wake it up if nothing else does.
810 16. A malformed maildirsize file could cause Exim to calculate negative values
811 for the mailbox size or file count. Odd effects could occur as a result.
812 The maildirsize information is now recalculated if the size or filecount
815 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
816 support for a long time. Removed HAVE_SYS_VFS_H.
818 18. Updated exipick to current release from John Jetmore.
820 19. Allow an empty sender to be matched against a lookup in an address list.
821 Previously the only cases considered were a regular expression, or an
824 20. Exim went into a mad DNS lookup loop when doing a callout where the
825 host was specified on the transport, if the DNS lookup yielded more than
828 21. The RFC2047 encoding function was originally intended for short strings
829 such as real names; it was not keeping to the 75-character limit for
830 encoded words that the RFC imposes. It now respects the limit, and
831 generates multiple encoded words if necessary. To be on the safe side, I
832 have increased the buffer size for the ${rfc2047: expansion operator from
835 22. Failure to deliver a bounce message always caused it to be frozen, even if
836 there was an errors_to setting on the router. The errors_to setting is now
839 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
840 canonical form (fully expanded) before being placed in
841 $sender_host_address.
843 24. Updated eximstats to version 1.33
845 25. Include certificate and key file names in error message when GnuTLS fails
846 to set them up, because the GnuTLS error message doesn't include the name
847 of the failing file when there is a problem reading it.
849 26. Expand error message when OpenSSL has problems setting up cert/key files.
852 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
853 (this can affect the format of dates).
855 28. exim_tidydb, when checking for the continued existence of a message for
856 which it has found a message-specific retry record, was not finding
857 messages that were in split spool directories. Consequently, it was
858 deleting retry records that should have stayed in existence.
860 29. eximstats updated to version 1.35
861 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
862 1.35 - bugfix such that pie charts by volume are generated correctly
864 30. The SPA authentication driver was not abandoning authentication and moving
865 on to the next authenticator when an expansion was forced to fail,
866 contradicting the general specification for all authenticators. Instead it
867 was generating a temporary error. It now behaves as specified.
869 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
870 (the order specifies the preference for clients). The order is now AES256,
871 AES128, 3DES, ARCFOUR128.
873 31. Small patch to Sieve code - explicitly set From: when generating an
876 32. Exim crashed if a remote delivery caused a very long error message to be
877 recorded - for instance if somebody sent an entire SpamAssassin report back
878 as a large number of 550 error lines. This bug was coincidentally fixed by
879 increasing the size of one of Exim's internal buffers (big_buffer) that
880 happened as part of the Exiscan merge. However, to be on the safe side, I
881 have made the code more robust (and fixed the comments that describe what
884 33. Some experimental protocols are using DNS PTR records for new purposes. The
885 keys for these records are domain names, not reversed IP addresses. The
886 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
887 leaves it alone. Component reversal etc. now happens only for IP addresses.
890 34. The host_aton() function is supposed to be passed a string that is known
891 to be a valid IP address. However, in the case of IPv6 addresses, it was
892 not checking this. This is a hostage to fortune. Exim now panics and dies
893 if the condition is not met. A case was found where this could be provoked
894 from a dnsdb PTR lookup with an IPv6 address that had more than 8
895 components; fortuitously, this particular loophole had already been fixed
896 by change 4.50/55 or 4.44/33 above.
898 If there are any other similar loopholes, the new check in host_aton()
899 itself should stop them being exploited. The report I received stated that
900 data on the command line could provoke the exploit when Exim was running as
901 exim, but did not say which command line option was involved. All I could
902 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
906 35. There was a buffer overflow vulnerability in the SPA authentication code
907 (which came originally from the Samba project). I have added a test to the
908 spa_base64_to_bits() function which I hope fixes it.
911 36. The daemon start-up calls getloadavg() while still root for those OS that
912 need the first call to be done as root, but it missed one case: when
913 deliver_queue_load_max is set with deliver_drop_privilege. This is
914 necessary for the benefit of the queue runner, because there is no re-exec
915 when deliver_drop_privilege is set.
917 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
918 was in use, was not putting the data itself into the right store pool;
919 consequently, it could be overwritten for a subsequent message in the same
920 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
923 38. Sometimes the final signoff response after QUIT could fail to get
924 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
925 < 0 before doing a fflush(). This bug looks as though it goes back to the
926 introduction of TLS in release 3.20, but "sometimes" must have been rare
927 because the tests only now provoked it.
933 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
934 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
935 Now it returns an integer. A number of calls were still expecting a T/F
936 return. Fortuitously, in all cases, the tests worked in OK situations,
937 which is the norm. However, things would have gone wrong on any write
938 failures on the smtp file descriptor. This function is used when sending
939 messages over SMTP and also when doing verify callouts.
941 2. When Exim is called to do synchronous delivery of a locally submitted
942 message (the -odf or -odi options), it no longer closes stderr before doing
945 3. Implemented the mua_wrapper option.
947 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
949 5. Implemented the functions header_remove(), header_testname(),
950 header_add_at_position(), and receive_remove_recipient(), and exported them
953 6. If an ACL "warn" statement specified the addition of headers, Exim already
954 inserted X-ACL-Warn: at the start if there was no header name. However, it
955 was not making this test for the second and subsequent header lines if
956 there were newlines in the string. This meant that an invalid header could
957 be inserted if Exim was badly configured.
959 7. Allow an ACL "warn" statement to add header lines at the start or after all
960 the Received: headers, as well as at the end.
962 8. Added the rcpt_4xx retry error code.
964 9. Added postmaster_mailfrom=xxx to callout verification option.
966 10. Added mailfrom=xxxx to the callout verification option, for verify=
969 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
970 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
972 12. Inserted some casts to stop certain compilers warning when using pointer
973 differences as field lengths or precisions in printf-type calls (mostly
974 affecting debugging statements).
976 13. Added optional readline() support for -be (dynamically loaded).
978 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
979 same clock tick as a message's arrival, so that its received time was the
980 same as the "first fail" time on the retry record, and that message
981 remained on the queue past the ultimate address timeout, every queue runner
982 would try a delivery (because it was past the ultimate address timeout) but
983 after another failure, the ultimate address timeout, which should have then
984 bounced the address, did not kick in. This was a "< instead of <=" error;
985 in most cases the first failure would have been in the next clock tick
986 after the received time, and all would be well.
988 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
989 being recognized when the domain list was tested by the match_domain
990 condition in an expansion string.
992 16. Added the ${str2b64: operator.
994 17. Exim was always calling setrlimit() to set a large limit for the number of
995 processes, without checking whether the existing limit was already
996 adequate. (It did check for the limit on file descriptors.) Furthermore,
997 errors from getrlimit() and setrlimit() were being ignored. Now they are
998 logged to the main and panic logs, but Exim does carry on, to try to do its
999 job under whatever limits there are.
1001 18. Imported PCRE 5.0.
1003 19. Trivial typo in log message " temporarily refused connection" (the leading
1006 20. If the log selector return_path_on_delivery was set and an address was
1007 redirected to /dev/null, the delivery process crashed because it assumed
1008 that a return path would always be set for a "successful" delivery. In this
1009 case, the whole delivery is bypassed as an optimization, and therefore no
1012 21. Internal re-arrangement: the function for sending a challenge and reading
1013 a response while authentication was assuming a zero-terminated challenge
1014 string. It's now changed to take a pointer and a length, to allow for
1015 binary data in such strings.
1017 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1019 23. Exim was not respecting finduser_retries when seeking the login of the
1020 uid under which it was called; it was always trying 10 times. (The default
1021 setting of finduser_retries is zero.) Also, it was sleeping after the final
1022 failure, which is pointless.
1024 24. Implemented tls_on_connect_ports.
1026 25. Implemented acl_smtp_predata.
1028 26. If the domain in control=submission is set empty, Exim assumes that the
1029 authenticated id is a complete email address when it generates From: or
1030 Sender: header lines.
1032 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1033 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1034 chown and chgrp in /bin and hostname in /usr/bin.
1036 28. Exim was keeping the "process log" file open after each use, just as it
1037 does for the main log. This opens the possibility of it remaining open for
1038 long periods when the USR1 signal hits a daemon. Occasional processlog
1039 errors were reported, that could have been caused by this. Anyway, it seems
1040 much more sensible not to leave this file open at all, so that is what now
1043 29. The long-running daemon process does not normally write to the log once it
1044 has entered its main loop, and it closes the log before doing so. This is
1045 so that log files can straightforwardly be renamed and moved. However,
1046 there are a couple of unusual error situations where the daemon does write
1047 log entries, and I had neglected to close the log afterwards.
1049 30. The text of an SMTP error response that was received during a remote
1050 delivery was being truncated at 512 bytes. This is too short for some of
1051 the long messages that one sometimes sees. I've increased the limit to
1054 31. It is now possible to make retry rules that apply only when a message has a
1055 specific sender, in particular, an empty sender.
1057 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1058 it possible to be selective about when SMTP synchronization is enforced.
1060 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1062 32. Implemented hosts_connection_nolog.
1064 33. Added an ACL for QUIT.
1066 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1069 35. Added mailbox_size and mailbox_filecount to appendfile.
1071 36. Added control = no_multiline_responses to ACLs.
1073 37. There was a bug in the logic of the code that waits for the clock to tick
1074 in the case where the clock went backwards by a substantial amount such
1075 that the microsecond fraction of "now" was more than the microsecond
1076 fraction of "then" (but the whole seconds number was less).
1078 38. Added support for the libradius Radius client library this is found on
1079 FreeBSD (previously only the radiusclient library was supported).
1085 1. When certain lookups returned multiple values in the form name=value, the
1086 quoting of the values was not always being done properly. Specifically:
1087 (a) If the value started with a double quote, but contained no whitespace,
1089 (b) If the value contained whitespace other than a space character (i.e.
1090 tabs or newlines or carriage returns) it was not quoted.
1091 This fix has been applied to the mysql and pgsql lookups by writing a
1092 separate quoting function and calling it from the lookup code. The fix
1093 should probably also be applied to nisplus, ibase and oracle lookups, but
1094 since I cannot test any of those, I have not disturbed their existing code.
1096 2. A hit in the callout cache for a specific address caused a log line with no
1097 reason for rejecting RCPT. Now it says "Previous (cached) callout
1098 verification failure".
1100 3. There was an off-by-one bug in the queryprogram router. An over-long
1101 return line was truncated at 256 instead of 255 characters, thereby
1102 overflowing its buffer with the terminating zero. As well as fixing this, I
1103 have increased the buffer size to 1024 (and made a note to document this).
1105 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1106 when Exim is waiting for an SMTP response from a remote server, Exim
1107 restarts its select() call on the socket, thereby resetting its timeout.
1108 This is not a problem when such interrupts are rare. Somebody set up a cron
1109 job to run exiwhat every 2 minutes, which is less than the normal select()
1110 timeout (5 or 10 minutes). This meant that the select() timeout never
1111 kicked in because it was always reset. I have fixed this by comparing the
1112 time when an interrupt arrives with the time at the start of the first call
1113 to select(). If more time than the timeout has elapsed, the interrupt is
1114 treated as a timeout.
1116 5. Some internal re-factoring in preparation for the addition of Sieve
1117 extensions (by MH). In particular, the "personal" test is moved to a
1118 separate function, and given an option for scanning Cc: and Bcc: (which is
1119 not set for Exim filters).
1121 6. When Exim created an email address using the login of the caller as the
1122 local part (e.g. when creating a From: or Sender: header line), it was not
1123 quoting the local part when it contained special characters such as @.
1125 7. Installed new OpenBSD configuration files.
1127 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1128 try to make them clearer.
1130 9. Callout options, other than the timeout value, were being ignored when
1131 verifying sender addresses in header lines. For example, when using
1133 verify = header_sender/callout=no_cache
1135 the cache was (incorrectly) being used.
1137 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1138 only the Cygwin environment.
1140 11. When return_path_on_delivery was set as a log selector, if different remote
1141 addresses in the same message used different return paths and parallel
1142 remote delivery occurred, the wrong values would sometimes be logged.
1143 (Whenever a remote delivery process finished, the return path value from
1144 the most recently started remote delivery process was logged.)
1146 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1147 lines when AUTH and/or TLS are in use. This is the "received protocol"
1148 field. Exim used to use "asmtp" for authenticated SMTP, without any
1149 indication (in the protocol name) for TLS use. Now it follows the RFC and
1150 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1151 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1152 names appear in log lines as well as in Received: header lines.
1154 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1155 extensions, and comparison tests, and to fix some bugs.
1157 14. Changes to the "personal" filter test:
1159 (1) The test was buggy in that it was just doing the equivalent of
1160 "contains" tests on header lines. For example, if a user's address was
1161 anne@some.where, the "personal" test would incorrectly be true for
1163 To: susanne@some.where
1165 This test is now done by extracting each address from the header in turn,
1166 and checking the entire address. Other tests that are part of "personal"
1167 are now done using regular expressions (for example, to check local parts
1168 of addresses in From: header lines).
1170 (2) The list of non-personal local parts in From: addresses has been
1171 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1172 taken from the Sieve specification recommendations.
1174 (3) If the message contains any header line starting with "List-" it is
1175 treated as non-personal.
1177 (4) The test for "circular" in the Subject: header line has been removed
1178 because it now seems ill-conceived.
1180 15. Minor typos in src/EDITME comments corrected.
1182 16. Installed latest exipick from John Jetmore.
1184 17. If headers_add on a router specified a text string that was too long for
1185 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1186 of string_sprintf() is now avoided.
1188 18. $message_body_size was not set (it was always zero) when running the DATA
1189 ACL and the local_scan() function.
1191 19. For the "mail" command in an Exim filter, no default was being set for
1192 the once_repeat time, causing a random time value to be used if "once" was
1193 specified. (If the value happened to be <= 0, no repeat happened.) The
1194 default is now 0s, meaning "never repeat". The "vacation" command was OK
1195 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1196 (I found it when inspecting the code).
1198 20. There is now an overall timeout for performing a callout verification. It
1199 defaults to 4 times the callout timeout, which applies to individual SMTP
1200 commands during the callout. The overall timeout applies when there is more
1201 than one host that can be tried. The timeout is checked before trying the
1202 next host. This prevents very long delays if there are a large number of
1203 hosts and all are timing out (e.g. when the network connections are timing
1204 out). The value of the overall timeout can be changed by specifying an
1205 additional sub-option for "callout", called "maxwait". For example:
1207 verify = sender/callout=5s,maxwait=20s
1209 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1210 to the end before writing, but this should make it even safer).
1212 22. Exim was forgetting that it had advertised PIPELINING for the second and
1213 subsequent messages on an SMTP connection. It was also not resetting its
1214 memory on STARTTLS and an internal HELO.
1216 23. When Exim logs an SMTP synchronization error within a session, it now
1217 records whether PIPELINING has been advertised or not.
1219 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1220 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1221 rather than long int.
1223 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1225 26. Added the never_mail option to autoreply.
1231 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1232 crash if the getsockname() call failed; this can happen if a connection is
1233 closed very soon after it is established. The problem was simply in the
1234 order in which certain operations were done, causing Exim to try to write
1235 to the SMTP stream before it had set up the file descriptor. The bug has
1236 been fixed by making things happen in the correct order.
1242 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1243 before the connection was closed, thus losing the rejection response.
1245 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1246 some early Solaris releases, but causes trouble in current releases where
1247 socklen_t is defined.
1249 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1252 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1255 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1256 that releasing the top of it at the end releases what was used for sub-
1257 expansions (unless the block got too big). However, discard this block if
1258 the first thing is a variable or header, so that we can use its block when
1259 it is dynamic (useful for very large $message_headers, for example).
1261 6. Lookups now cache *every* query, not just the most recent. A new, separate
1262 store pool is used for this. It can be recovered when all lookup caches are
1263 flushed. Lookups now release memory at the end of their result strings.
1264 This has involved some general refactoring of the lookup sources.
1266 7. Some code has been added to the store_xxx() functions to reduce the amount
1267 of flapping under certain conditions.
1269 8. log_incoming_interface used to affect only the <= reception log lines. Now
1270 it causes the local interface and port to be added to several more SMTP log
1271 lines, for example "SMTP connection from", and rejection lines.
1273 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1275 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1277 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1278 could be overwritten at the end of the current message (or the start of a
1279 new message if it was set in a HELO ACL). The value is now preserved for
1280 the duration of the SMTP connection.
1282 12. If a transport had a headers_rewrite setting, and a matching header line
1283 contained an unqualified address, that address was qualified, even if it
1284 did not match any rewriting rules. The underlying bug was that the values
1285 of the flags that permit the existence of unqualified sender and recipient
1286 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1287 non-local messages, and by -bnq for local messages) were not being
1288 preserved with the message after it was received.
1290 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1291 "next input=" as part of the text comprising the host identity instead of
1292 the correct text. The code was using the same buffer for two different
1293 strings. However, depending on which order the printing function evaluated
1294 its arguments, the bug did not always show up. Under Linux, for example, my
1295 test suite worked just fine.
1297 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1298 doesn't work with some older versions of Perl. It has been changed to "my",
1299 which in any case is probably the better facility to use.
1301 15. A really picky compiler found some instances of statements for creating
1302 error messages that either had too many or two few arguments for the format
1305 16. The size of the buffer for calls to the DNS resolver has been increased
1306 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1307 for addresses that have a lot of PTR records. This alleviates a problem; it
1308 does not fully solve it.
1310 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1311 buffer now truncates the list and logs the incident, which is the same
1312 action as happens when Exim is looking up a host name and its aliases.
1313 Previously in this situation something unpredictable would happen;
1314 sometimes it was "internal error: store_reset failed".
1316 18. If a server dropped the connection unexpectedly when an Exim client was
1317 using GnuTLS and trying to read a response, the client delivery process
1318 crashed while trying to generate an error log message.
1320 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1321 string, for example:
1323 warn message = H1: something\nH2: something
1325 the text was added as a single header line from Exim's point of view
1326 though it ended up OK in the delivered message. However, searching for the
1327 second and subsequent header lines using $h_h2: did not work. This has been
1328 fixed. Similarly, if a system filter added multiple headers in this way,
1329 the routers could not see them.
1331 20. Expanded the error message when iplsearch is called with an invalid key to
1332 suggest using net-iplsearch in a host list.
1334 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1335 ACLs are no longer actually imposed (and a message to that effect is
1338 22. If a "gecos" field in a passwd entry contained escaped characters, in
1339 particular, if it contained a \" sequence, Exim got it wrong when building
1340 a From: or a Sender: header from that name. A second bug also caused
1341 incorrect handling when an unquoted " was present following a character
1342 that needed quoting.
1344 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1345 was not being matched caselessly.
1347 24. Arranged for all hyphens in the exim.8 source to be escaped with
1350 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1351 properly. Recipient callout cache records were still being keyed to include
1352 the sender, even when use_sender was set false. This led to far more
1353 callouts that were necessary. The sender is no longer included in the key
1354 when use_sender is false.
1356 26. Added "control = submission" modifier to ACLs.
1358 27. Added the ${base62d: operator to decode base 62 numbers.
1360 28. dnsdb lookups can now access SRV records.
1362 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1363 the configuration file.
1365 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1366 (-v) mode. This makes the output for a verbose queue run more intelligible.
1368 31. Added a use_postmaster feature to recipient callouts.
1370 32. Added the $body_zerocount variable, containing the number of binary zero
1371 bytes in the message body.
1373 33. The time of last modification of the "new" subdirectory is now used as the
1374 "mailbox time last read" when there is a quota error for a maildir
1377 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1379 35. Added +ignore_unknown as a special item in host lists.
1381 36. Code for decoding IPv6 addresses in host lists is now included, even if
1382 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1383 address was recognized as an IP address, but was then not correctly decoded
1384 into binary, causing unexpected and incorrect effects when compared with
1391 1. Very minor rewording of debugging text in manualroute to say "list of
1392 hosts" instead of "hostlist".
1394 2. If verify=header_syntax was set, and a header line with an unqualified
1395 address (no domain) and a large number of spaces between the end of the
1396 name and the colon was received, the reception process suffered a buffer
1397 overflow, and (when I tested it) crashed. This was caused by some obsolete
1398 code that should have been removed. The fix is to remove it!
1400 3. When running in the test harness, delay a bit after writing a bounce
1401 message to get a bit more predictability in the log output.
1403 4. Added a call to search_tidyup() just before forking a reception process. In
1404 theory, someone could use a lookup in the expansion of smtp_accept_max_
1405 per_host which, without the tidyup, could leave open a database connection.
1407 5. Added the variables $recipient_data and $sender_data which get set from a
1408 lookup success in an ACL "recipients" or "senders" condition, or a router
1409 "senders" option, similar to $domain_data and $local_part_data.
1411 6. Moved the writing of debug_print from before to after the "senders" test
1414 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1415 problems for message scanning, either using a data ACL, or using
1416 local_scan() because the Received: header was not generated till after they
1417 were called (in order to set the time as the time of reception completion).
1418 I have revised the way this works. The header is now generated after the
1419 body is received, but before the ACL or local_scan() are called. After they
1420 are run, the timestamp in the header is updated.
1426 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1427 before starting a queue runner without re-exec. This happened only when
1428 deliver_drop_privilege was set or when the Exim user was set to root. The
1429 effect of the bug was that timeouts during subsequent deliveries caused
1430 crashes instead of being properly handled. The handler is now left at its
1431 default (and expected) setting.
1433 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1434 message, again when deliver_drop_privilege is set or Exim is run as root.
1435 The bug described in (1) was not present in this case, but the tidying up
1436 of the other signals was missing. I have made the two cases consistent.
1438 3. The ignore_target_hosts setting on a manualroute router was being ignored
1439 for hosts that were looked up using the /MX notation.
1441 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1444 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1445 operated on the sender address. After changing the $sender_address to <>
1446 for the sender address verify, Exim was re-instated it as the original
1447 (before rewriting) address, but remembering that it had rewritten it, so it
1448 wasn't rewriting it again. This bug also had the effect of breaking the
1449 sender address verification caching when the sender address was rewritten.
1451 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1452 This has been changed so that if the ip literal address matches
1453 ignore_target_hosts, the router declines.
1455 7. Added expansion conditions match_domain, match_address, and match_local_
1456 part (NOT match_host).
1458 8. The placeholder for the Received: header didn't have a length field set.
1460 9. Added code to Exim itself and to exim_lock to test for a specific race
1461 condition that could lead to file corruption when using MBX delivery. The
1462 issue is with the lockfile that is created in /tmp. If this file is removed
1463 after a process has opened it but before that process has acquired a lock,
1464 there is the potential for a second process to recreate the file and also
1465 acquire a lock. This could lead to two Exim processes writing to the file
1466 at the same time. The added code performs the same test as UW imapd; it
1467 checks after acquiring the lock that its file descriptor still refers to
1468 the same named file.
1470 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1471 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1474 11. Added the smtp_active_hostname option. If used, this will typically be made
1475 to depend on the incoming interface address. Because $interface_address is
1476 not set up until the daemon has forked a reception process, error responses
1477 that can happen earlier (such as "too many connections") no longer contain
1480 12. If an expansion in a condition on a "warn" statement fails because a lookup
1481 defers, the "warn" statement is abandoned, and the next ACL statement is
1482 processed. Previously this caused the whole ACL to be aborted.
1484 13. Added the iplsearch lookup type.
1486 14. Added ident_timeout as a log selector.
1488 15. Added tls_certificate_verified as a log selector.
1490 16. Added a global option tls_require_ciphers (compare the smtp transport
1491 option of the same name). This controls incoming TLS connections.
1493 17. I finally figured out how to make tls_require_ciphers do a similar thing
1494 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1495 before starting the TLS session.
1497 18. Tabs are now shown as \t in -bP output.
1499 19. If the log selector return_path_on_delivery was set, Exim crashed when
1500 bouncing a message because it had too many Received: header lines.
1502 20. If two routers both had headers_remove settings, and the first one included
1503 a superfluous trailing colon, the final name in the first list and the
1504 first name in the second list were incorrectly joined into one item (with a
1505 colon in the middle).
1511 1. Added -C and -D options to the exinext utility, mainly to make it easier
1512 to include in the automated testing, but these could be helpful when
1513 multiple configurations are in use.
1515 2. The exinext utility was not formatting the output nicely when there was
1516 an alternate port involved in the retry record key, nor when there was a
1517 message id as well (for retries that were specific to a specific message
1518 and a specific host). It was also confused by IPv6 addresses, because of
1519 the additional colons they contain. I have fixed the IPv4 problem, and
1520 patched it up to do a reasonable job for IPv6.
1522 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1523 delivery, the log line now contains "pipelined" if PIPELINING was used.
1525 4. An SMTP transport process used to panic and die if the bind() call to set
1526 an explicit outgoing interface failed. This has been changed; it is now
1527 treated in the same way as a connect() failure.
1529 5. A reference to $sender_host_name in the part of a conditional expansion
1530 that was being skipped was still causing a DNS lookup. This no longer
1533 6. The def: expansion condition was not recognizing references to header lines
1534 that used bh_ and bheader_.
1536 7. Added the _cache feature to named lists.
1538 8. The code for checking quota_filecount in the appendfile transport was
1539 allowing one more file than it should have been.
1541 9. For compatibility with Sendmail, the command line option
1549 and sets the incoming protocol and host name (for trusted callers). The
1550 host name and its colon can be omitted when only the protocol is to be set.
1551 Note the Exim already has two private options, -pd and -ps, that refer to
1552 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1553 "s", but I don't think that's a major issue.
1555 10. A number of refactoring changes to the code, none of which should affect
1558 (a) The number of logging options was getting close to filling up the
1559 32-bit word that was used as a bit map. I have split them into two classes:
1560 those that are passed in the argument to log_write(), and those that are
1561 only ever tested independently outside of that function. These are now in
1562 separate 32-bit words, so there is plenty of room for expansion again.
1563 There is no change in the user interface or the logging behaviour.
1565 (b) When building, for example, log lines, the code previously used a
1566 macro that called string_cat() twice, in order to add two strings. This is
1567 not really sufficiently general. Furthermore, there was one instance where
1568 it was actually wrong because one of the argument was used twice, and in
1569 one call a function was used. (As it happened, calling the function twice
1570 did not affect the overall behaviour.) The macro has been replaced by a
1571 function that can join an arbitrary number of extra strings onto a growing
1574 (c) The code for expansion conditions now uses a table and a binary chop
1575 instead of a serial search (which was left over from when there were very
1576 few conditions). Also, it now recognizes conditions like "pam" even when
1577 the relevant support is not compiled in: a suitably worded error message is
1578 given if an attempt is made to use such a condition.
1580 11. Added ${time_interval:xxxxx}.
1582 12. A bug was causing one of the ddress fields not to be passed back correctly
1583 from remote delivery subprocesses. The field in question was not being
1584 subsequently used, so this caused to problems in practice.
1586 13. Added new log selectors queue_time and deliver_time.
1588 14. Might have fixed a bug in maildirsizefile handling that threw up
1589 "unexpected character" debug warnings, and recalculated the data
1590 unnecessarily. In any case, I expanded the warning message to give more
1593 15. Added the message "Restricted characters in address" to the statements in
1594 the default ACL that block characters like @ and % in local parts.
1596 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1597 Three changes have been made:
1599 (a) There was a serious bug; a negative response to MAIL caused the whole
1600 recipient domain to be cached as invalid, thereby blocking all messages
1601 to all local parts at the same domain, from all senders. This bug has
1602 been fixed. The domain is no longer cached after a negative response to
1603 MAIL if the sender used is not empty.
1605 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1608 (c) A new callout option, "use_sender" has been added for people who want
1609 the modified behaviour.
1615 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1618 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1619 indeed breaks things for older releases.
1621 3. Added additional logging to the case where there is a problem reading data
1622 from a filter that is running in a subprocess using a pipe, in order to
1623 try to track down a specific problem.
1625 4. Testing facility fudge: when running in the test harness and attempting
1626 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1627 getting "No route to host". Convert this to a timeout.
1629 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1632 6. Some OS don't have socklen_t but use size_t instead. This affects the
1633 fifth argument of getsockopt() amongst other things. This is now
1634 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1635 can be set for individual OS. I have set it for SunOS5, OSF1, and
1636 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1637 some earlier ones do not.
1639 7. Change 4.30/15 was not doing the test caselessly.
1641 8. The standard form for an IPv6 address literal was being rejected by address
1642 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1643 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1644 this, as well as the form without the "IPv6" on the front (but only when
1645 address literals are enabled, of course).
1647 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1649 10. Exim crashed if a message with an empty sender address specified by -f
1650 encountered a router with an errors_to setting. This could be provoked only
1651 by a command such as
1655 where an empty string was supplied; "<>" did not hit this bug.
1657 11. Installed PCRE release 4.5.
1659 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1660 remained set. It is now erased.
1662 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1663 times from message ids (which are base 36 rather than the normal 62).
1665 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1666 were being counted as actual protocol errors, and logged if the log
1667 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1668 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1669 and DATA following a set of rejected RCPTs do not count as protocol errors.
1670 In other words, Exim assumes they were pipelined, though this may not
1671 actually be the case. Of course, in all cases the client gets an
1672 appropriate error code.
1674 15. If a lookup fails in an ACL condition, a message about the failure may
1675 be available; it is used if testing the ACL cannot continue, because most
1676 such messages specify what the cause of the deferral is. However, some
1677 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1678 that caused an old message to be retained and used if a later statement
1679 caused a defer, replacing the real cause of the deferral.
1681 16. If an IP address had so many PTR records that the DNS lookup buffer
1682 was not large enough to hold them, Exim could crash while trying to process
1683 the truncated data. It now detects and logs this case.
1685 17. Further to 4.21/58, another change has been made: if (and only if) the
1686 first line of a message (the first header line) ends with CRLF, a bare LF
1687 in a subsequent header line has a space inserted after it, so as not to
1688 terminate the header.
1690 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1691 unnecessarily, used atoi() instead of strtol(), and didn't check the
1692 termination when getting file sizes from file names by regex.
1694 19. Completely re-implemented the support for maildirsize files, in the light
1695 of a number of problems with the previous contributed implementation
1696 (4.30/29). In particular:
1698 . If the quota is zero, the maildirsize file is maintained, but no quota is
1701 . If the maildir directory does not exist, it is created before any attempt
1702 to write a maildirsize file.
1704 . The quota value in the file is just a cache; if the quota is changed in
1705 the transport, the new value overrides.
1707 . A regular expression is available for excluding directories from the
1710 20. The autoreply transport checks the characters in options that define the
1711 message's headers; it allows continued headers, but it was checking with
1712 isspace() after an embedded newline instead of explicitly looking for a
1715 21. If all the "regular" hosts to which an address was routed had passed their
1716 expiry times, and had not reached their retry times, the address was
1717 bounced, even if fallback hosts were defined. Now Exim should go on to try
1720 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1721 equivalent code in the SMTP transport. Some hosts send humungous responses
1722 to HELO/EHLO, more than 1024 it seems.
1724 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1725 gives compiler warnings in some environments. I've now done it "properly",
1728 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1729 (because of problems with the built-in one) was declared to return uschar *
1730 instead of char *, causing compiler failure.
1732 25. Fixed a file descriptor leak when processing alias/forward files.
1734 26. Fixed a minor format string issue in dbfn.c.
1736 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1738 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1739 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1742 29. When a sender address is verified, it is cached, to save repeating the test
1743 when there is more than one recipient in a message. However, when the
1744 verification involves a callout, it is possible for different callout
1745 options to be set for different recipients. It is too complicated to keep
1746 track of this in the cache, so now Exim always runs a verification when a
1747 callout is required, relying on the callout cache for the optimization.
1748 The overhead is duplication of the address routing, but this should not be
1751 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1752 to be verified with callout=postmaster, and the main callout worked but the
1753 postmaster check failed, the verification correctly failed. However, if a
1754 subsequent RCPT command asked for sender verification *without* the
1755 postmaster check, incorrect caching caused this verification also to fail,
1758 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1759 it was not caching the DNS options (qualify_single, search_parents) that
1760 were used when the lookup failed. A subsequent lookup with different
1761 options therefore always gave the same answer, though there were cases
1762 where it should not have. (Example: a "domains = !$mx_any" option on a
1763 dnslookup router: the "domains" option is always processed without any
1764 widening, but the router might have qualify_single set.) Now Exim uses the
1765 cached value only when the same options are set.
1767 32. Added John Jetmore's "exipick" utility to the distribution.
1769 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1770 than a timeout (e.g. a certificate is required, and is not provided), an
1771 Exim server now closes the connection immediately. Previously it waited for
1772 the client to close - but if the client is SSL, it seems that they each
1773 wait for each other, leading to a delay before one of them times out.
1775 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1776 maintained 0.8.x compatibility because I don't think many are using it, and
1777 it is clearly obsolete.
1779 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1782 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1783 client certificate was expired. A simple patch fixes this, though I don't
1784 understand the full logic of why the verify callback is called multiple
1787 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1788 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1789 which causes problems with some clients (such as the Certicom SSL Plus
1790 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1791 disables the coutermeasure allowing Eudora to connect."
1793 38. Exim was not checking that a write() to a log file succeeded. This could
1794 lead to Bad Things if a log got too big, in particular if it hit a file
1795 size limit. Exim now panics and dies if it cannot write to a log file, just
1796 as it does if it cannot open a log file.
1798 39. Modified OS/Makefile-Linux so that it now contains
1800 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1802 The two -D definitions ensure that Exim is compiled with large file
1803 support, which makes it possible to handle log files that are bigger than
1806 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1807 instance) a domain was checked against a named list that involved a lookup,
1808 causing $domain_data to be set, then another domain was checked against the
1809 same list, then the first domain was re-checked, the value of $domain_data
1810 after the final check could be wrong. In particular, if the second check
1811 failed, it could be set empty. This bug probably also applied to
1814 41. The strip_trailing_dot option was not being applied to the address given
1815 with the -f command-line option.
1817 42. The code for reading a message's header from the spool was incrementing
1818 $received_count, but never initializing it. This meant that the value was
1819 incorrect (doubled) while delivering a message in the same process in which
1820 it was received. In the most common configuration of Exim, this never
1821 happens - a fresh exec is done - but it can happen when
1822 deliver_drop_privilege is set.
1824 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1825 it now includes up to 150 characters of the unexpected data in the log
1828 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1829 and building data strings. The size of both of these buffers was 10 000
1830 bytes - far larger than anybody would *ever* want, thought I. Needless to
1831 say, somebody hit the limit. I have increased the maximum line length to
1832 20 000 and the maximum data length of concatenated lines to 100 000. I have
1833 also fixed two bugs, because there was no checking on these buffers. Tsk,
1834 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1835 buffer is too small.
1837 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1838 lsearch lookups. Now it does.
1840 46. When parsing a route_list item in a manualroute router, a fixed-length
1841 buffer was used for the list of hosts. I made this 1024 bytes long,
1842 thinking that nobody would ever have a list of hosts that long. Wrong.
1843 Somebody had a whole pile of complicated expansion conditions, and the
1844 string was silently truncated, leading to an expansion error. It turns out
1845 that it is easier to change to an unlimited length (owing to other changes
1846 that have happened since this code was originally written) than to build
1847 structure for giving a limitation error. The length of the item that
1848 expands into the list of hosts is now unlimited.
1850 47. The lsearch lookup could not handle data where the length of text line was
1851 more than 4095 characters. Such lines were truncated, leading to shortened
1852 data being returned. It should now handle lines of any length.
1854 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1855 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1858 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1859 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1860 no longer happens in comment lines. A list of replacements is now placed
1861 at the head of all of the source files, except those whose only change is
1862 to replace PERL_COMMAND in the very first #! line.
1864 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1865 messages on the queue, with a bottom-up merge sort, using code contributed
1866 by Michael Haardt. This should make operations like -bp somewhat faster on
1867 large queues. It won't affect queue runners, except when queue_run_in_order
1870 51. Installed eximstats 1.31 in the distribution.
1872 52. Added support for SRV lookups to the dnslookup router.
1874 53. If an ACL referred to $message_body or $message_body_end, the value was not
1875 reset for any messages that followed in the same SMTP session.
1877 54. The store-handling optimization for building very long strings was not
1878 differentiating between the different store pools. I don't think this
1879 actually made any difference in practice, but I've tidied it.
1881 55. While running the routers to verify a sender address, $sender_address
1882 was still set to the sender address. This is wrong, because when routing to
1883 send a bounce to the sender, it would be empty. Therefore, I have changed
1884 it so that, while verifying a sender address, $sender_address is set to <>.
1885 (There is no change to what happens when verifying a recipient address.)
1887 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1888 target A or AAAA records (if not already returned) without resetting the
1889 qualify_single or search_parents options of the DNS resolver. These are
1890 inappropriate in this case because the targets of MX and SRV records must
1891 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1892 target that, when qualified, matched something in the local domain. These
1893 two options are now turned off when doing these lookups.
1895 57. It seems that at least some releases of Reiserfs (which does not have the
1896 concept of a fixed number of inodes) returns zero and not -1 for the
1897 number of available inodes. This interacted badly with check_spool_inodes,
1898 which assumed that -1 was the "no such thing" setting. What I have done is
1899 to check that the total number of inodes is greater than zero before doing
1900 the test of how many are available.
1902 58. When a "warn" ACL statement has a log_message modifier, the message is
1903 remembered, and not repeated. This is to avoid a lot of repetition when a
1904 message has many recipients that cause the same warning to be written.
1905 Howewer, Exim was preserving the list of already written lines for an
1906 entire SMTP session, which doesn't seem right. The memory is now reset if a
1907 new message is started.
1909 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1910 debugging output unless log_rewrite was also set.
1912 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1913 of (int)(handle) when we know that handle contains (void *)(-1).
1915 61. The Exim daemon panic-logs an error return when it closes the incoming
1916 connection. However "connection reset by peer" seems to be common, and
1917 isn't really an error worthy of noting specially, so that particular error
1920 62. When Exim is trying to find all the local interfaces, it used to panic and
1921 die if the ioctl to get the interface flags failed. However, it seems that
1922 on at least one OS (Solaris 9) it is possible to have an interface that is
1923 included in the list of interfaces, but for which you get a failure error
1924 for this call. This happens when the interface is not "plumbed" into a
1925 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1926 failure of the "get flags" call assumes that the interface is down.
1928 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1929 makes life easier for people who are doing arithmetic on fields extracted
1930 from dates, where you often get leading zeros that should not be
1931 interpreted as octal.
1933 64. Added qualify_domain to the redirect router, to override the global
1936 65. If a pathologically long header line contained very many addresses (the
1937 report of this problem mentioned 10 000) and each of them was rewritten,
1938 Exim could use up a very large amount of memory. (It kept on making new
1939 copies of the header line as it rewrote, and never released the old ones.)
1940 At the expense of a bit more processing, the header rewriting function has
1941 been changed so that it no longer eats memory in this way.
1943 66. The generation of the Received: header has been moved from the time that a
1944 message starts to be received, to the time that it finishes. The timestamp
1945 in the Received: header should now be very close to that of the <= log
1946 line. There are two side-effects of this change:
1948 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1949 logged header lines no longer include the local Received: line, because
1950 it has not yet been created. The same applies to a copy of the message
1951 that is returned to a non-SMTP sender when a message is rejected.
1953 (b) When a filter file is tested using -bf, no additional Received: header
1954 is added to the test message. After some thought, I decided that this
1957 This change does not affect the value of $received_for. It is still set
1958 after address rewriting, but before local_scan() is called.
1960 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1962 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1963 gave an unhelpful panic error message, and a defer error. I have managed to
1964 change this behaviour so that it now rejects any supplied certificate,
1965 which seems right, as the list of acceptable certificates is empty.
1967 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1968 gave an unhelpful defer error. I have not managed to make this reject any
1969 supplied certificates, but the error message it gives is "no certificate
1970 supplied", which is not helpful.
1972 70. exigrep's output now also includes lines that are not associated with any
1973 message, but which match the given pattern. Implemented by a patch from
1974 Martin Sluka, which also tidied up the Perl a bit.
1976 71. Recipient callout verification, like sender verification, was using <> in
1977 the MAIL FROM command. This isn't really the right thing, since the actual
1978 sender may affect whether the remote host accepts the recipient or not. I
1979 have changed it to use the actual sender in the callout; this means that
1980 the cache record is now keyed on a recipient/sender pair, not just the
1981 recipient address. There doesn't seem to be a real danger of callout loops,
1982 since a callout by the remote host to check the sender would use <>.
1983 [SEE ABOVE: changed after hitting problems.]
1985 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1986 temporary errors. However, in the case of such a code being given after
1987 the end of a data transmission (i.e. after ".") Exim was failing to write
1988 a retry record for the message. (Yes, there was some broken host that was
1989 actually sending 8xx at this point.)
1991 73. An unknown lookup type in a host list could cause Exim to panic-die when
1992 the list was checked. (An example that provoked this was putting <; in the
1993 middle of a list instead of at the start.) If this happened during a DATA
1994 ACL check, a -D file could be left lying around. This kind of configuration
1995 error no longer causes Exim to die; instead it causes a defer errror. The
1996 incident is still logged to the main and panic logs.
1998 74. Buglet left over from Exim 3 conversion. The message "too many messages
1999 in one connection" was written to the rejectlog but not the mainlog, except
2000 when address rewriting (yes!) was being logged.
2002 75. Added write_rejectlog option.
2004 76. When a system filter was run not as root (that is, when system_filter_user
2005 was set), the values of the $n variables were not being returned to the
2006 main process; thus, they were not subsequently available in the $sn
2009 77. Added +return_path_on_delivery log selector.
2011 78. A connection timeout was being treated differently from recipients deferred
2012 when testing hosts_max_try with a message that was older than the host's
2013 retry timeout. (The host should not be counted, thus allowing all hosts to
2014 be tried at least once before bouncing.) This may have been the cause of an
2015 occasionally reported bug whereby a message would remain on the queue
2016 longer than the retry timeout, but would be bounced if a delivery was
2017 forced. I say "may" because I never totally pinned down the problem;
2018 setting up timeout/retry tests is difficult. See also the next item.
2020 79. The ultimate address timeout was not being applied to errors that involved
2021 a combination of host plus message (for example, a timeout on a MAIL
2022 command). When an address resolved to a number of possible hosts, and they
2023 were not all tried for each delivery (e.g. because of hosts_max_try), a
2024 message could remain on the queue longer than the retry timeout.
2026 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2029 81. Fixed an obscure SMTP outgoing bug which required at least the following
2030 conditions: (a) there was another message waiting for the same server;
2031 (b) the server returned 5xx to all RCPT commands in the first message so
2032 that the message was not completed; (c) the server dropped the connection
2033 or gave a negative response to the RSET that Exim sends to abort the
2034 transaction. The observed case was a dropped connection after DATA that had
2035 been sent in pipelining mode. That is, the server had advertised PIPELINING
2036 but was not implementing it correctly. The effect of the bug was incorrect
2037 behaviour, such as trying another host, and this could lead to a crash.
2043 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2044 and daemon.c were passed as pointers to ints; they should have been
2045 pointers to socklen_t variables (which are typically unsigned ints).
2047 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2050 3. Fixed a really odd bug that affected only the testing scheme; patching a
2051 certain fixed string in the binary changed the value of another string that
2052 happened to be identical to the end of the original first string.
2054 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2055 name", it returns that address as the IP address. On some operating
2056 systems (e.g. Solaris), it also passes back the IP address string as the
2057 "host name". However, on others (e.g. Linux), it passes back an empty
2058 string. Exim wasn't checking for this, and was changing the host name to an
2059 empty string, assuming it had been canonicized.
2061 5. Although rare, it is permitted to have more than one PTR record for a given
2062 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2063 all the names associated with an address, because they do in Solaris.
2064 However, it seems that they do not in Linux for data that comes from the
2065 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2066 I found this out when I moved to a new Linux workstation and tried to run
2067 the Exim test suite.
2069 To get round this problem I have changed the code so that it now does its
2070 own call to the DNS to look up PTR records when searching for a host name.
2071 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2072 addresses that are only in /etc/hosts are still found.
2074 This behaviour is, however, controlled by an option called host_lookup_
2075 order, which defaults to "bydns:byaddr". If people want to use the other
2076 order, or indeed, just use one or the other means of lookup, they can
2077 specify it in this variable.
2079 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2080 some operating systems, this comes back from gethostbyaddr() as an empty
2081 string, and this is what Exim used to test for. However, it seems that in
2082 other systems, "." is yielded. Exim now tests for this case too.
2084 7. The values of check_spool_space and check_log_space are now held internally
2085 as a number of kilobytes instead of an absolute number of bytes. If a
2086 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2087 kilobyte. This means that much larger values can be stored.
2089 8. Exim monitor: an attempt to get the action menu when not actually pointing
2090 at a message produces an empty menu entitled "No message selected". This
2091 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2092 no entries in it ("Shell widget menu has zero width and/or height"). So I
2093 have added a single, blank menu entry in this case.
2095 9. Added ${quote_local_part.
2097 10. MIME decoding is now applied to the contents of Subject: header lines when
2100 11. Now that a reference to $sender_host_address automatically causes a reverse
2101 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2102 host lookup before query-style lookups in lists that might use this
2103 variable. This has therefore been abolished, and the "net-" prefix is no
2104 longer necessary for query-style lookups.
2106 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2107 have been a typo for LFLAGS, so it has been changed.
2109 13. The install script calls Exim with "-C /dev/null" in order to find the
2110 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2111 to be output. Howeve, since Exim outputs its version number before the
2112 error, it didn't break the script. It just looked ugly. I fixed this by
2113 always allowing "-C /dev/null" if the caller is root.
2115 14. Ignore overlarge ACL variable number when reading spool file - insurance
2116 against a later release with more variables having written the file.
2118 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2119 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2120 this, as well as the form without the "IPv6" on the front.
2122 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2123 OS/Makefile-Darwin file.
2125 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2126 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2128 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2129 variable. However, some OS use ints for inodes. I've added cast to long int
2130 to get rid of the compiler warning.
2132 19. I had forgotten to lock out "/../" in configuration file names when
2133 ALT_CONFIG_PREFIX was set.
2135 20. Routers used for verification do not need to specify transports. However,
2136 if such a router generated a host list, and callout was configured, Exim
2137 crashed, because it could not find a port number from the (non-existent)
2138 transport. It now assumes port 25 in this circumstance.
2140 21. Added the -t option to exigrep.
2142 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2143 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2144 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2145 them is set, LOOKUP_LSEARCH is forced.
2147 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2149 24. Added sender and host information to the "rejected by local_scan()" log
2150 line; previously there was no indication of these.
2152 25. Added .include_if_exists.
2154 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2155 Linux. It turns out that not all file systems support this. Apparently some
2156 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2157 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2158 not supported on the file descriptor, is now ignored when Exim is trying to
2159 sync a directory. This applies only to Linux.
2161 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2163 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2164 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2167 29. Added support for maildirsize files from supplied patch (modified a bit).
2169 30. The use of :fail: followed by an empty string could lead Exim to respond to
2170 sender verification failures with (e.g.):
2172 550 Verification failed for <xxx>
2173 550 Sender verify failed
2175 where the first response line was missing the '-' that indicates it is not
2176 the final line of the response.
2178 31. The loop for finding the name of the user that called Exim had a hardwired
2179 limit of 10; it now uses the value of finduser_retries, which is used for
2180 all other user lookups.
2182 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2185 33. Exim was neglecting to zero errno before one call of strtol() when
2186 expanding a string and expecting an integer value. On some systems this
2187 resulted in spurious "integer overflow" errors. Also, it was casting the
2188 result into an int without checking.
2190 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2191 did not work. The code looks as if it has *never* worked, though it appears
2192 to have been documented since at least releast 1.62. I have made it work.
2194 35. The "timeout_DNS" error in retry rules, also documented since at least
2195 1.62, also never worked. As it isn't clear exactly what this means, and
2196 clearly it isn't a major issue, I have abolished the feature by treating it
2197 as "timeout", and writing a warning to the main and panic logs.
2199 36. The display of retry rules for -brt wasn't always showing the error code
2202 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2203 timeout_connect_A, timeout_connect_MX.
2205 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2206 to the empty sender.
2208 39. The daemon was not analysing the content of -oX till after it had closed
2209 stderr and disconnected from the controlling terminal. This meant that any
2210 syntax errors were only noted on the panic log, and the return code from
2211 the command was 0. By re-arranging the code a little, I've made the
2212 decoding happen first, so such errors now appear on stderr, and the return
2213 code is 1. However, the actual setting up of the sockets still happens in
2214 the disconnected process, so errors there are still only recorded on the
2217 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2218 connections (as happens on some IP stacks) was logged at start up time as
2219 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2220 it from "IPv6 and IPv4", which means that two separate sockets are being
2223 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2224 says whether AF_INET or AF_INET6 was passed as an argument.
2226 42. Exiwhat output was messed up when time zones were included in log
2229 43. Exiwhat now gives more information about the daemon's listening ports,
2230 and whether -tls-on-connect was used.
2232 44. The "port" option of the smtp transport is now expanded.
2234 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2235 silently ignored. Now an error message is written to the main and panic
2238 46. There's a new ACL modifier called "logwrite" which writes to a log file
2239 as soon as it is encountered.
2241 47. Added $local_user_uid and $local_user_gid at routing time.
2243 48. Exim crashed when trying to verify a sender address that was being
2246 49. Exim was recognizing only a space character after ".include". It now also
2247 recognizes a tab character.
2249 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2250 extracting the relevant information from the specification. The man page no
2251 longer contains scrambled data for the -d option, and I've added a section
2252 at the front about calling Exim under different names.
2254 51. Added "extra_headers" argument to the "mail" command in filter files.
2256 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2259 53. Installed eximstats 1.29.
2261 54. Added transport_filter_timeout as a generic transport option.
2263 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2264 Cc: header lines. This was required by RFC 822, but it not required by RFC
2267 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2268 incoming messages that did not have them. Now it does so only if the
2269 message originates locally, that is, if there is no associated remote host
2270 address. When Resent- header lines are present, this applies to the Resent-
2271 lines rather than the non-Resent- lines.
2273 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2274 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2276 58. Messages for configuration errors now include the name of the main
2277 configuration file - useful now that there may be more than one file in a
2278 list (.included file names were always shown).
2280 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2281 for those rare installations that do not start the daemon as root or run it
2282 setuid root. I've cut out the call to initgroups() if the daemon is not
2285 60. The Exim user and group can now be bound into the binary as text strings
2286 that are looked up at the start of Exim's processing.
2288 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2290 62. Added $mailstore_basename variable.
2292 63. Installed patch to sieve.c from Michael Haardt.
2294 64. When Exim failed to open the panic log after failing to open the main log,
2295 the original message it was trying to log was written to stderr and debug
2296 output, but if they were not available (the usual case in production), it
2297 was lost. Now it is written to syslog before the two lines that record the
2298 failures to open the logs.
2300 65. Users' Exim filters run in subprocesses under the user's uid. It is
2301 possible for a "deliver" command or an alias in a "personal" command to
2302 provoke an address rewrite. If logging of address rewriting is configured,
2303 this fails because the process is not running as root or exim. There may be
2304 a better way of dealing with this, but for the moment (because 4.30 needs
2305 to be released), I have disabled address rewrite logging when running a
2306 filter in a non-root, non-exim process.
2312 1. The buildconfig auxiliary program wasn't quoting the value set for
2313 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2314 not defined. This bug was masked in 4.22 by the effect that was fixed in
2317 2. Some messages that were rejected after a message id was allocated were
2318 shown as "incomplete" by exigrep. It no longer does this for messages that
2319 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2321 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2322 have allow_domain_literals set, the ID did not get logged in the <= line.
2323 Domain literals are now always recognized in Message-ID: header lines.
2325 4. The first argument for a ${extract expansion item is the key name or field
2326 number. Leading and trailing spaces in this item were not being ignored,
2327 causing some misleading effects.
2329 5. When deliver_drop_privilege was set, single queue runner processes started
2330 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2331 same command in the process it spins off) were not dropping privilege.
2333 6. When the daemon running as "exim" started a queue runner, it always
2334 re-executed Exim in the spun-off process. This is a waste of effort when
2335 deliver_drop_privilege is set. The new process now just calls the
2336 queue-runner function directly.
2342 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2345 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2346 ignored. Though the use of -oP was forcing the writing of a pid file, it
2347 was always written to the default place.
2349 3. If the message "no IP address found for host xxxx" is generated during
2350 incoming verification, it is now followed by identification of the incoming
2351 connection (so you can more easily find what provoked it).
2353 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2355 5. Added some features to "harden" Exim a bit more against certain attacks:
2357 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2358 be put in Local/Makefile. This is like the never_users runtime option,
2359 but it cannot be overridden. The default setting is "root".
2361 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2362 prefix string with which any file named in a -C command line option
2365 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2366 is retained for -C and -D only if the caller of Exim is root. Without
2367 it, the exim user may also use -C and -D and retain privilege.
2369 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2370 command line option is disabled.
2372 6. Macro names set by the -D option must start with an upper case letter, just
2373 like macro names defined in the configuration file.
2375 7. Added "dereference=" facility to LDAP.
2377 8. Two instances of the typo "uknown" in the source files are fixed.
2379 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2380 the Configure-Makefile script screwed up while processing it.
2382 10. Incorporated PCRE 4.4.
2384 11. The SMTP synchronization check was not operating right at the start of an
2385 SMTP session. For example, it could not catch a HELO sent before the client
2386 waited for the greeting. There is now a check for outstanding input at the
2387 point when the greeting is written. Because of the duplex, asynchronous
2388 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2389 way, but not yet received, when the check is performed.
2391 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2392 on TCP/IP sockets, because this apparently causes some broken clients to
2395 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2396 unchanged) from the Cygwin maintainer.
2398 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2399 was supported instead of testing for mbx. Effectively a typo.
2401 15. The spa authenticator server code was not checking that the input it
2402 received was valid base64.
2404 16. The debug output line for the "set" modifier in ACLs was not showing the
2405 name of the variable that was being set.
2407 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2409 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2410 lookup on its own. Something else was needed to trigger the lookup. For
2411 example, a match in host_lookup or the need for a host name in a host list.
2412 Now, if $sender_host_name is referenced and the host name has not yet been
2413 looked up, a lookup is performed. If the lookup fails, the variable remains
2414 empty, and $host_lookup_failed is set to "1".
2416 19. Added "eqi" as a case-independent comparison operator.
2418 20. The saslauthd authentication condition could segfault if neither service
2419 nor realm was specified.
2421 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2422 error message that was logged was misleading, and incoming SMTP
2423 connections were dropped. The message is now more accurate, and temporary
2424 errors are given to SMTP connections.
2426 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2427 (except RSET) with a 421 error, until QUIT is received. However, it was
2428 failing to send a response to QUIT.
2430 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2431 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2432 ACL is now run after the helo_try_verify_hosts code.
2434 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2435 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2436 case-independent, so other case variants are also recognized). Apparently
2437 some systems use these upper case variants.
2439 25. If more than two messages were waiting for the same host, and a transport
2440 filter was specified for the transport, Exim sent two messages over the
2441 same TCP/IP connection, and then failed with "socket operation on non-
2442 socket" when it tried to send the third.
2444 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2446 27. The extern definition of crypt16() in expand.c was not being excluded when
2447 the OS had its own crypt16() function.
2449 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2450 as a preferred synonym for return_size_limit, both as an option and as an
2453 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2455 30. Changed the default configuration ACL to relax the local part checking rule
2456 for addresses that are not in any local domains. For these addresses,
2457 slashes and pipe symbols are allowed within local parts, but the sequence
2458 /../ is explicitly forbidden.
2460 31. SPA server authentication was not clearing the challenge buffer before
2463 32. log_message in a "warn" ACL statement was writing to the reject log as
2464 well as to the main log, which contradicts the documentation and doesn't
2465 seem right (because no rejection is happening). So I have stopped it.
2467 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2468 However, I am unable to do any testing of this.
2470 34. Fixed an infelicity in the appendfile transport. When checking directories
2471 for a mailbox, to see if any needed to be created, it was accidentally
2472 using path names with one or more superfluous leading slashes; tracing
2473 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2475 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2476 contents are added to the log line that is written for every discarded
2477 recipient. (Previously a log_message setting was ignored.)
2479 36. The ${quote: operator now quotes the string if it is empty.
2481 37. The install script runs exim in order to find its version number. If for
2482 some reason other than non-existence or emptiness, which it checks, it
2483 could not run './exim', it was installing it with an empty version number,
2484 i.e. as "exim-". This error state is now caught, and the installation is
2487 38. An argument was missing from the function that creates an error message
2488 when Exim fails to connect to the socket for saslauthd authentication.
2489 This could cause Exim to crash, or give a corrupted message.
2491 39. Added isip, isip4, and isip6 to ${if conditions.
2493 40. The ACL variables $acl_xx are now saved with the message, and can be
2494 accessed later in routers, transports, and filters.
2496 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2497 strings in the file are not string-expanded.
2499 42. If a MAIL command specified a SIZE value that was too large to fit into an
2500 int variable, the check against message_size_limit failed. Such values are
2501 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2502 day this will have to be increased, but I don't think I want to be around
2503 when emails are that large.
2510 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2511 iconv() is not standard in FreeBSD.
2513 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2514 IPv6 enabled. The observed symptom was a segmentation fault on return from
2515 the function os_common_find_running_interfaces() in src/os.c.
2517 3. In the check_special_case() function in daemon.c I had used "errno" as an
2518 argument name, which causes warnings on some systems. This was basically a
2519 typo, since it was named "eno" in the comments!
2521 4. The code that waits for the clock to tick (at a resolution of some fraction
2522 of a second) so as to ensure message-id uniqueness was always waiting for
2523 at least one whole tick, when it could have waited for less. [This is
2524 almost certainly not relevant at current processor speeds, where it is
2525 unlikely to ever wait at all. But we try to future-proof.]
2527 5. The function that sleeps for a time interval that includes fractions of a
2528 second contained a race. It did not block SIGALRM between setting the
2529 timer, and suspending (a couple of lines later). If the interval was short
2530 and the sigsuspend() was delayed until after it had expired, the suspension
2531 never ended. On busy systems this could lead to processes getting stuck for
2534 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2535 process, before it forks any delivery processes. The open lookup caching
2536 mechanism meant that the open file or database connection was passed into
2537 the delivery process. The problem was that delivery processes always tidy
2538 up cached lookup data. This could cause a problem for the next delivery
2539 process started by the queue runner, because the external queue runner
2540 process does not know about the closure. So the next delivery process
2541 still has data in the lookup cache. In the case of a file lookup, there was
2542 no problem because closing a file descriptor in a subprocess doesn't affect
2543 the parent. However, if the lookup was caching a connection to a database,
2544 the connection was closed, and the second delivery process was likely to
2545 see errors such as "PGSQL: query failed: server closed the connection
2546 unexpectedly". The problem has been fixed by closing all cached lookups
2547 in a queue runner before running a delivery process.
2549 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2550 seem to have the "const" qualifier which it has on other OS. I've
2553 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2554 *of the same type* (client or server) with the same public name that an
2555 error should be diagnosed.
2557 9. When Exim looked up a host name for an IP address, but failed to find the
2558 original IP address when looking up the host name (a safety check), it
2559 output the message "<ip address> does not match any IP for NULL", which was
2560 confusing, to say the least. The bug was that the host name should have
2561 appeared instead of "NULL".
2563 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2564 user that is built into the binary, and the -C or -D options is used, root
2565 privilege is dropped before the configuration file is read. In addition,
2566 logging is switched to stderr instead of the normal log files. If the
2567 configuration then re-defines the Exim user, the unprivileged environment
2568 is probably not what is expected, so Exim logs a panic warning message (but
2571 However, if deliver_drop_privilege is set, the unprivileged state may well
2572 be exactly what is intended, so the warning has been cut out in that case,
2573 and Exim is allowed to try to write to its normal log files.
2579 1. smtp_return_error_details was not giving details for temporary sender
2580 or receiver verification errors.
2582 2. Diagnose a configuration error if two authenticators have the same public
2585 3. Exim used not to create the message log file for a message until the first
2586 delivery attempt. This could be confusing when incoming messages were held
2587 for policy or load reasons. The message log file is now created at the time
2588 the message is received, and an initial "Received" line is written to it.
2590 4. The automatically generated man page for command line options had a minor
2591 bug that caused no ill effects; however, a more serious problem was that
2592 the procedure for building the man page automatically didn't always
2593 operate. Consequently, release 4.20 contains an out-of-date version. This
2594 shouldn't happen again.
2596 5. When building Exim with embedded Perl support, the script that builds the
2597 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2598 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2600 6. The freeze_tell option was not being used for messages that were frozen on
2601 arrival, either by an ACL or by local_scan().
2603 7. Added the smtp_incomplete_transaction log selector.
2605 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2606 was accepting AUTH without a new EHLO.
2608 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2609 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2612 10. Logging of TCP/IP connections (when configured) now happens in the main
2613 daemon process instead of the child process, so that the TCP/IP connection
2614 count is more accurate (but it can never be perfect).
2616 11. The use of "drop" in a nested ACL was not being handled correctly in the
2617 outer ACL. Now, if condition failure induced by the nested "drop" causes
2618 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2619 or "require"), the connection is dropped.
2621 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2622 that yield "discard" can now be used with an "accept" or a "discard" verb,
2623 but an error is generated for any others (because I can't see a useful way
2624 to define what should happen).
2626 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2627 are now processed in the same way as lines in the Exim configuration file.
2628 In particular, continuation lines are supported.
2630 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2632 15. Added -ti meaning -t -i.
2634 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2635 domains, and warn by logging if others are found.
2637 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2638 returned from GIFCONF when Exim is trying to find the list of interfaces on
2639 a host. The code in os.c has been modified to copy each ifreq to an aligned
2640 structure in all cases.
2642 Also, in some cases, the returned ifreq's were being copied to a 'struct
2643 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2644 means the last couple of bytes of an IPv6 address could be chopped if the
2645 ifreq contained only a normal sockaddr (14 bytes storage).
2647 18. Named domain lists were not supported in the hosts_treat_as_local option.
2648 An entry such as +xxxx was not recognized, and was treated as a literal
2651 19. Ensure that header lines added by a DATA ACL are included in the reject log
2652 if the ACL subsequently rejects the message.
2654 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2655 MD5 (which is deprecated).
2657 21. When testing a filter file using -bf, Exim was writing a message when it
2658 took the sender from a "From " line in the message, but it was not doing so
2659 when it took $return_path from a Return-Path: header line. It now does.
2661 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2662 with a valid header line field name (a series of printing characters
2663 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2665 23. Changed "disc" in the source to "disk" to conform to the documentation and
2666 the book and for uniformity.
2668 24. Ignore Sendmail's -Ooption=value command line item.
2670 25. When execve() failed while trying to run a command in a pipe transport,
2671 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2672 could be confused with a return value of 69 from the command itself. This
2673 has been changed to 127, the value the shell returns if it is asked to run
2674 a non-existent command. The wording for the related log line suggests a
2675 non-existent command as the problem.
2677 26. If received_header_text expands to an empty string, do not add a Received:
2678 header line to the message. (Well, it adds a token one on the spool, but
2679 marks it "old" so that it doesn't get used or transmitted.)
2681 27. Installed eximstats 1.28 (addition of -nt option).
2683 28. There was no check for failure on the call to getsockname() in the daemon
2684 code. This can fail if there is a shortage of resources on the system, with
2685 ENOMEM, for example. A temporary error is now given on failure.
2687 29. Contrary to the C standard, it seems that in some environments, the
2688 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2689 program. Exim now does this explicitly; it affects the formatting of
2690 timestamps using strftime().
2692 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2693 complaints. I've now initialized all the variables, to avoid this.
2695 32. Header lines added by a system filter were not being "seen" during
2696 transport-time rewrites.
2698 33. The info_callback() function passed to OpenSSL is set up with type void
2699 (*)(SSL *, int, int), as described somewhere. However, when calling the
2700 function (actually a macro) that sets it up, the type void(*)() is
2701 expected. I've put in a cast to prevent warnings from picky compilers.
2703 34. If a DNS black list lookup found a CNAME record, but there were no A
2704 records associated with the domain it pointed at, Exim crashed.
2706 35. If a DNS black list lookup returned more than one A record, Exim ignored
2707 all but the first. It now scans all returned addresses if a particular IP
2708 value is being sought. In this situation, the contents of the
2709 $dnslist_value variable are a list of all the addresses, separated by a
2712 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2713 to accept a host name and all its aliases if the forward lookup for any of
2714 them yielded the IP address of the incoming connection. Now it accepts only
2715 those names whose forward lookup yields the correct IP address. Any other
2716 names are discarded. This closes a loophole whereby a rogue DNS
2717 administrator could create reverse DNS records to break through a
2718 wildcarded host restriction in an ACL.
2720 37. If a user filter or a system filter that ran in a subprocess used any of
2721 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2722 the wrong values were passed to the pipe command ($thisaddress had the
2723 value of $0, $0 had the value of $1, etc). This bug was introduced by
2724 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2726 38. Improved the line breaking for long SMTP error messages from ACLs.
2727 Previously, if there was no break point between 40 and 75 characters, Exim
2728 left the rest of the message alone. Two changes have been made: (a) I've
2729 reduced the minimum length to 35 characters; (b) if it can't find a break
2730 point between 35 and 75 characters, it looks ahead and uses the first one
2731 that it finds. This may give the occasional overlong line, but at least the
2732 remaining text gets split now.
2734 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2735 file descriptors might be low, and that setting 1000 would always raise it.
2736 It turns out that in some environments, the limit is already over 1000 and
2737 that lowering it causes trouble. So now Exim takes care not to decrease it.
2739 40. When delivering a message, the value of $return_path is set to $sender_
2740 address at the start of routing (routers may change the value). By an
2741 oversight, this default was not being set up when an address was tested by
2742 -bt or -bv, which affected the outcome if any router or filter referred to
2745 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2746 log only when "message" or "log_message" are set. However, if one of the
2747 conditions was an address verification, or a call to a nested ACL, the
2748 messages generated by the underlying test were being passed through. This
2749 no longer happens. The underlying message is available in $acl_verify_
2750 message for both "message" and "log_message" expansions, so it can be
2751 passed through if needed.
2753 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2754 new expansion $bh_ to give the encoded byte string without charset
2755 translation. Translation happens only if iconv() is available; HAVE_ICONV
2756 indicates this at build time. HEADERS_CHARSET gives the charset to
2757 translate to; headers_charset can change it in the configuration, and
2758 "headers charset" can change it in an individual filter file.
2760 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2761 that creates RFC 2047 encoded "words" labels them as that charset instead
2762 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2763 expansion operator; (ii) when Exim creates a From: line for a local
2764 message; (iii) when a header line is rewritten to include a "phrase" part.
2766 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2767 buggy, causing it to skip the first lines of messages whose message ID
2768 ended in 'D'. This would not have bitten before Exim release 4.14, because
2769 message IDs were unlikely to end in 'D' before then. The effect was to have
2770 incorrect size information for certain domains.
2772 45. #include "config.h" was missing at the start of the crypt16.c module. This
2773 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2776 46. If there was a timeout during a "random" callout check, Exim treated it as
2777 a failure of the random address, and carried on sending RSET and the real
2778 address. If the delay was just some slowness somewhere, the response to the
2779 original RCPT would be taken as a response to RSET and so on, causing
2780 mayhem of various kinds.
2782 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2783 when I implemented it. It didn't allow for the fact that some option values
2784 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2785 the right test for positive values.
2787 48. Domain names in DNS records are case-independent. Exim always looks them up
2788 in lower case. Some resolvers return domain names in exactly the case they
2789 appear in the zone file, that is, they may contain uppercase letters. Not
2790 all resolvers do this - some return always lower case. Exim was treating a
2791 change of case by a resolver as a change of domain, similar to a widening
2792 of a domain abbreviation. This triggered its re-routing code and so it was
2793 trying to route what was effectively the same domain again. This normally
2794 caused routing to fail (because the router wouldn't handle the domain
2795 twice). Now Exim checks for this case specially, and just changes the
2796 casing of the domain that it ultimately uses when it transmits the message
2799 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2802 50. If a filter generated a file delivery with a non-absolute name (possible if
2803 no home directory exists for the router), the forbid_file option was not
2806 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2807 the existing equality matching.
2809 52. Exim was using ints instead of ino_t variables in some places where it was
2810 dealing with inode numbers.
2812 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2813 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2814 called TMPDIR, and if it finds it is different, it changes its value.
2816 54. The smtp_printf() function is now made available to local_scan() so
2817 additional output lines can be written before returning. There is also an
2818 smtp_fflush() function to enable the detection of a dropped connection.
2819 The variables smtp_input and smtp_batched_input are exported to
2822 55. Changed the default runtime configuration: the message "Unknown user"
2823 has been removed from the ACL, and instead placed on the localuser router,
2824 using the cannot_route_message feature. This means that any verification
2825 failures that generate their own messages won't get overridden. Similarly,
2826 the "Unrouteable address" message that was in the ACL for unverifiable
2827 relay addresses has also been removed.
2829 56. Added hosts_avoid_esmtp to the smtp transport.
2831 57. The exicyclog script was not checking for the esoteric option
2832 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2833 will work only if exicyclog is run under the appropriate euid.
2835 58. Following a discussion on the list, the rules by which Exim recognises line
2836 endings on incoming messages have been changed. The -dropcr and drop_cr
2837 options are now no-ops, retained only for backwards compatibility. The
2838 following line terminators are recognized: LF CRLF CR. However, special
2839 processing applies to CR:
2841 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2842 nor a local message in the state where . is a terminator.
2844 (ii) If a bare CR is encountered in a header line, an extra space is added
2845 after the line terminator so as not to end the header. The reasoning
2846 behind this is that bare CRs in header lines are most likely either
2847 to be mistakes, or people trying to play silly games.
2849 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2850 was being incorrectly given as 18 bytes larger than it should have been.
2851 This is a VOB (very old bug).
2853 60. This may never have affected anything current, but just in case it has:
2854 When the local host is found other than at the start of a list of hosts,
2855 the local host, those with the same MX, and any that follow, are discarded.
2856 When the list in question was part of a longer list of hosts, the following
2857 hosts (not currently being processed) were also being discarded. This no
2858 longer happens. I'm not sure if this situation could ever has previously
2861 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2864 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2865 header. This is something that is recommended in a new Internet Draft, and
2866 is something that is documented as being done by Sendmail. There are two
2867 possible values. For messages generated by the autoreply transport, Exim
2870 Auto-Submitted: auto-replied
2872 whereas for all other generated messages (e.g. bounces) it adds
2874 Auto-Submitted: auto-generated
2876 63. The "personal" condition in filters now includes a test for the
2877 Auto-Submitted: header. If it contains the string "auto-" the message it
2878 not considered personal.
2880 64. Added rcpt_include_affixes as a generic transport option.
2882 65. Added queue_only_override (default true).
2884 66. Added the syslog_duplication option.
2886 67. If what should have been the first header line of a message consisted of
2887 a space followed by a colon, Exim was mis-interpreting it as a header line.
2888 It isn't of course - it is syntactically invalid and should therefore be
2889 treated as the start of the message body. The misbehaviour could have
2890 caused a number of strange effects, including loss of data in subsequent
2891 header lines, and spool format errors.
2893 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2894 client host had authenticated. This control can now be exercised by an ACL
2895 for more flexibility.
2897 69. By default, callouts do not happen when testing with -bh. There is now a
2898 variant, -bhc, which does actually run the callout code, including
2899 consulting and updating the callout cache.
2901 70. Added support for saslauthd authentication, courtesy of Alexander
2904 71. If statvfs() failed on the spool or log directories while checking their
2905 size for availability, Exim confusingly gave the error "space shortage".
2906 Furthermore, in debugging mode it crashed with a floating point exception.
2907 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2908 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2909 serious problem, Exim now writes to the main and panic logs when this
2910 happens, with details of the failure. It then refuses to accept the
2911 incoming message, giving the message "spool directory problem" or "log
2912 directory problem" with a 421 code for SMTP messages.
2914 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2915 0, 1, and 2 exist, because some OS complain for execs without them (see
2916 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2917 descriptors. However, the code omitted to check that the open succeeded,
2918 causing mysterious errors if for some reason the permissions on /dev/null
2919 got screwed. Now Exim writes a message to the main and panic logs, and
2920 bombs out if it can't open /dev/null.
2922 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2923 interact so that it is all more flexible. It is supposed to remain
2924 backwards compatible. Also added extra_local_interfaces.
2926 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2927 to bomb out with an assertion failure - to the client this appears as a
2928 connection drop. This problem occurs in the part of the code that was taken
2929 from the Samba project. Fortunately, the assertion is in a very simple
2930 function, so I have fixed this by reproducing the function inline in the
2931 one place where it is called, and arranging for authentication to fail
2932 instead of killing the process with assert().
2934 75. The SPA client code was not working when the server requested OEM rather
2935 than Unicode encoding.
2937 76. Added code to make require_files with a specific uid setting more usable in
2938 the case where statting the file as root fails - usually a non-root-mounted
2939 NFS file system. When this happens and the failure is EACCES, Exim now
2940 forks a subprocess and does the per-uid checking as the relevant uid.
2942 77. Added process_log_path.
2944 78. If log_file_path was not explicitly set, a setting of check_log_space or
2945 check_log_inodes was ignored.
2947 79. If a space check for the spool or log partitions fails, the incident is now
2948 logged. Of course, in the latter case the data may get lost...
2950 80. Added the %p formatting code to string_format() so that it can be used to
2951 print addresses in debug_print(). Adjusted all the address printing in the
2952 debugging in store.c to use %p rather than %d.
2954 81. There was a concern that a line of code in smtp_in.c could overflow a
2955 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2956 initially expressed, the concern was not well-founded, because trailing
2957 spaces are removed early. However, if the trailing spaces were followed by
2958 a NULL, they did not get removed, so the overflow was possible. Two fixes
2961 (a) I re-wrote the offending code in a cleaner fashion.
2962 (b) If an incoming SMTP command contains a NULL character, it is rejected
2965 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2966 runs initgroups(), so that if the Exim user is in any additional groups,
2967 they will be used during message reception.
2973 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob