1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.150 2005/06/07 15:20:55 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.
64 SC/01 Eximstats: added -xls and the ability to specify output files
65 (patch written by Frank Heydlauf).
67 SC/02 Eximstats: use FileHandles for outputing results.
69 SC/03 Eximstats: allow any combination of xls, txt, and html output.
71 SC/04 Eximstats: fixed display of large numbers with -nvr option
73 SC/05 Eximstats: fixed merging of reports with empty tables.
75 SC/06 Eximstats: added the -include_original_destination flag
77 SC/07 Eximstats: removed tabs and trailing whitespace.
79 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
81 TK/06 MBOX spool code: Add real "From " MBOX separator line
82 so the .eml file is really in mbox format (even though
83 most programs do not really care). Patch from Alex Miller.
85 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
86 The latter is generated from $received_to and is only set if the
87 message has one envelope recipient. SA can use these headers,
88 obviously out-of-the-box. Patch from Alex Miller.
90 PH/08 The ${def test on a variable was returning false if the variable's
91 value was "0", contrary to what the specification has always said!
92 The result should be true unless the variable is empty.
94 PH/09 The syntax error of a character other than { following "${if
95 def:variable_name" (after optional whitespace) was not being diagnosed.
96 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
97 accidental colon was present, for example, could give incorrect results.
99 PH/10 Tidied the code in a number of places where the st_size field of a stat()
100 result is used (not including appendfile, where other changes are about
103 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
104 This involved changing a lot of size variables from int to off_t. It
105 should work with maildirs and everything.
111 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
112 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
114 TK/02 Fix ACL "control" statement not being available in MIME ACL.
116 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
118 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
119 to test Sieve filters that use "vacation".
121 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
122 that changes the way the GnuTLS parameters are stored in the cache file.
123 The new format can be generated externally. For backward compatibility,
124 if the data in the cache doesn't make sense, Exim assumes it has read an
125 old-format file, and it generates new data and writes a new file. This
126 means that you can't go back to an older release without removing the
129 PH/03 A redirect router that has both "unseen" and "one_time" set does not
130 work if there are any delivery delays because "one_time" forces the
131 parent to be marked "delivered", so its unseen clone is never tried
132 again. For this reason, Exim now forbids the simultaneous setting of
135 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
136 redirected to themselves ("homonym" addresses). Read the long ChangeLog
137 entry if you want to know the details. The fix, however, neglected to
138 consider the case when local delivery batching is involved. The test for
139 "previously delivered" was not happening when checking to see if an
140 address could be batched with a previous (undelivered) one; under
141 certain circumstances this could lead to multiple deliveries to the same
144 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
145 in its include files, and this causes problems building Exim.
147 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
148 header_syntax) but Exim was just ignoring anything given after a slash.
149 In particular, this caused confusion with an attempt to use "verify =
150 reverse_host_lookup/defer_ok". An error is now given when options are
151 supplied for verify items that do not have them. (Maybe reverse_host_
152 lookup should have a defer_ok option, but that's a different point.)
154 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
155 defined by RFC 821) to 2048, because there were problems with some AUTH
156 commands, and RFC 1869 says the size should be increased for extended
157 SMTP commands that take arguments.
159 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
162 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
163 "unknown" error; now it says that the functionality isn't in the binary.
165 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
166 an address' error message when a string expansion fails (syntax or
167 whatever). Otherwise the password may appear in the log. Following change
168 PH/42 below, there is no longer a chance of it appearing in a bounce
171 PH/11 Installed exipick version 20050225.0 from John Jetmore.
173 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
174 of its addresses was ever tried. (Bugzilla bug #2.)
176 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
177 the result incorrectly in the debug output. (It correctly added a newline
178 to what was transported.)
180 TF/01 Added $received_time.
182 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
183 commented out examples of how to interface to a virus scanner and to
184 SpamAssassin. Also added commented examples of av_scanner and
185 spamd_address settings.
187 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
188 and controls are allowed in which ACLs. There were a couple of minor
189 errors. Some of the entries in the conditions table (which is a table of
190 where they are NOT allowed) were getting very unwieldy; rewrote them as a
191 negation of where the condition IS allowed.
193 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
195 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
196 header file does not have a version number, so I've had to invent a new
197 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
198 API. The code is untested by me (my Linux distribution still has 0.3.2 of
199 radiusclient), but it was contributed by a Radius user.
201 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
202 files or directories, for OpenSSL.
204 PH/19 When an Exim process that is running as root has to create an Exim log
205 file, it does so in a subprocess that runs as exim:exim so as to get the
206 ownership right at creation (otherwise, other Exim processes might see
207 the file with the wrong ownership). There was no test for failure of this
208 fork() call, which would lead to the process getting stuck as it waited
209 for a non-existent subprocess. Forks do occasionally fail when resources
210 run out. I reviewed all the other calls to fork(); they all seem to check
213 PH/20 When checking for unexpected SMTP input at connect time (before writing
214 the banner), Exim was not dealing correctly with a non-positive return
215 from the read() function. If the client had disconnected by this time,
216 the result was a log entry for a synchronization error with an empty
217 string after "input=" when read() returned zero. If read() returned -1
218 (an event I could not check), uninitialized data bytes were printed.
219 There were reports of junk text (parts of files, etc) appearing after
222 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
224 PH/22 Added support for macro redefinition, and (re)definition in between
225 driver and ACL definitions.
227 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
228 forgetting to use the resulting value; it was using the unexpanded value.
230 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
231 hadn't been configured. The fix is from Juergen Kreileder, who
232 understands it better than I do:
234 "Here's what I see happening with three configured cyrus_sasl
235 authenticators configured (plain, login, cram-md5):
237 On startup auth_cyrus_sasl_init() gets called for each of these.
238 This means three calls to sasl_listmech() without a specified mech_list.
239 => SASL tests which mechs of all available mechs actually work
240 => three warnings about OTP not working
241 => the returned list contains: plain, login, cram-md5, digest-md5, ...
243 With the patch, sasl_listmech() also gets called three times. But now
244 SASL's mech_list option is set to the server_mech specified in the the
245 authenticator. Or in other words, the answer from sasl_listmech()
246 gets limited to just the mech you're testing for (which is different
248 => the return list contains just 'plain' or 'login', 'cram-md5' or
249 nothing depending on the value of ob->server_mech.
251 I've just tested the patch: Authentication still works fine,
252 unavailable mechs specified in the exim configuration are still
253 caught, and the auth.log warnings about OTP are gone."
255 PH/25 When debugging is enabled, the contents of the command line are added
256 to the debugging output, even when log_selector=+arguments is not
259 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
260 answer is "GNU", and only if the return is "GNU/something" is the answer
263 PH/27 $acl_verify_message is now set immediately after the failure of a
264 verification in an ACL, and so is available in subsequent modifiers. In
265 particular, the message can be preserved by coding like this:
267 warn !verify = sender
268 set acl_m0 = $acl_verify_message
270 Previously, $acl_verify_message was set only while expanding "message"
271 and "log_message" when a very denied access.
273 PH/28 Modified OS/os.c-Linux with
275 -#ifndef OS_LOAD_AVERAGE
276 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
278 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
279 nomenclature these days.)
281 PH/29 Installed patch from the Sieve maintainer that adds the options
282 sieve_useraddress and sieve_subaddress to the redirect router.
284 PH/30 In these circumstances:
285 . Two addresses routed to the same list of hosts;
286 . First host does not offer TLS;
287 . First host accepts first address;
288 . First host gives temporary error to second address;
289 . Second host offers TLS and a TLS session is established;
290 . Second host accepts second address.
291 Exim incorrectly logged both deliveries with the TLS parameters (cipher
292 and peerdn, if requested) that were in fact used only for the second
295 PH/31 When doing a callout as part of verifying an address, Exim was not paying
296 attention to any local part prefix or suffix that was matched by the
297 router that accepted the address. It now behaves in the same way as it
298 does for delivery: the affixes are removed from the local part unless
299 rcpt_include_affixes is set on the transport.
301 PH/32 Add the sender address, as F=<...>, to the log line when logging a
302 timeout during the DATA phase of an incoming message.
304 PH/33 Sieve envelope tests were broken for match types other than :is. I have
305 applied a patch sanctioned by the Sieve maintainer.
307 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
308 the uid or gid is negative. A case of a negative gid caused this to be
309 noticed. The fix allows for either to be negative.
311 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
312 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
315 PH/36 The change PH/12 above was broken. Fixed it.
317 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
318 the grounds that routing the same address twice would always produce the
319 same answer. This might have been true once, but it is certainly no
320 longer true now. Routing a child address may depend on the previous
321 routing that produced that child. Some complicated redirection strategies
322 went wrong when messages had multiple recipients, and made Exim's
323 behaviour dependent on the order in which the addresses were given.
325 I have moved the duplicate checking until after the routing is complete.
326 Exim scans the addresses that are assigned to local and remote
327 transports, and removes any duplicates. This means that more work will be
328 done, as duplicates will always all be routed, but duplicates are
329 presumably rare, so I don't expect this is of any significance.
331 For deliveries to pipes, files, and autoreplies, the duplicate checking
332 still happens during the routing process, since they are not going to be
335 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
336 It corrects a timeout issue with spamd. This is Ian's comment: "The
337 background is that sometimes spamd either never reads data from a
338 connection it has accepted, or it never writes response data. The exiscan
339 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
340 blindly assumes that writes won't block so it may never time out."
342 PH/39 Allow G after quota size as well as K and M.
344 PH/40 The value set for $authenticated_id in an authenticator may not contain
345 binary zeroes or newlines because the value is written to log lines and
346 to spool files. There was no check on this. Now the value is run through
347 the string_printing() function so that such characters are converted to
348 printable escape sequences.
350 PH/41 $message_linecount is a new variable that contains the total number of
351 lines in the message. Compare $body_linecount, which is the count for the
354 PH/42 Exim no longer gives details of delivery errors for specific addresses in
355 bounce and delay warning messages, except in certain special cases, which
358 (a) An SMTP error message from a remote host;
359 (b) A message specified in a :fail: redirection;
360 (c) A message specified in a "fail" command in a system filter;
361 (d) A message specified in a FAIL return from the queryprogram router;
362 (e) A message specified by the cannot_route_message router option.
364 In these cases only, Exim does include the error details in bounce and
365 warning messages. There are also a few cases where bland messages such
366 as "unrouteable address" or "local delivery error" are given.
368 PH/43 $value is now also set for the "else" part of a ${run expansion.
370 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
371 being worked on, but at least Exim now implements the latest version to
374 PH/45 In a pipe transport, although a timeout while waiting for the pipe
375 process to complete was treated as a delivery failure, a timeout while
376 writing the message to the pipe was logged, but erroneously treated as a
377 successful delivery. Such timeouts include transport filter timeouts. For
378 consistency with the overall process timeout, these timeouts are now
379 treated as errors, giving rise to delivery failures by default. However,
380 there is now a new Boolean option for the pipe transport called
381 timeout_defer, which, if set TRUE, converts the failures into defers for
382 both kinds of timeout. A transport filter timeout is now identified in
385 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
386 systems where "make" and "gmake" are different, calling "gmake" at top
387 level broke things. I've arranged for the value of $(MAKE) to be passed
388 from the Makefile to this script so that it can call the same version of
392 A note about Exim versions 4.44 and 4.50
393 ----------------------------------------
395 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
396 changes of various kinds. As a consequence, a big documentation update was
397 needed. This delayed the release for rather longer than seemed good, especially
398 in the light of a couple of (minor) security issues. Therefore, the changes
399 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
400 release. So 4.44 and 4.50 are in effect two different branches that both start
403 I have left the 4.50 change log unchanged; it contains all the changes since
404 4.43. The change log for 4.44 is below; many of its items are identical to
405 those for 4.50. This seems to be the most sensible way to preserve the
406 historical information.
412 1. Minor wording change to the doc/README.SIEVE file.
414 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
415 computation of the current number of files was incorrect.
417 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
418 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
421 4. Give more explanation in the error message when the command for a transport
422 filter fails to execute.
424 5. There are several places where Exim runs a non-Exim command in a
425 subprocess. The SIGUSR1 signal should be disabled for these processes. This
426 was being done only for the command run by the queryprogram router. It is
427 now done for all such subprocesses. The other cases are: ${run, transport
428 filters, and the commands run by the lmtp and pipe transports.
430 6. Added CONFIGURE_GROUP build-time option.
432 7. Some older OS have a limit of 256 on the maximum number of file
433 descriptors. Exim was using setrlimit() to set 1000 as a large value
434 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
435 systems. I've change it so that if it can't get 1000, it tries for 256.
437 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
438 was an oversight, and furthermore, ever since the addition of extra
439 controls (e.g. 4.43/32), the checks on when to allow different forms of
440 "control" were broken. There should now be diagnostics for all cases when a
441 control that does not make sense is encountered.
443 9. Added the /retain_sender option to "control=submission".
445 10. $recipients is now available in the predata ACL (oversight).
447 11. Tidy the search cache before the fork to do a delivery from a message
448 received from the command line. Otherwise the child will trigger a lookup
449 failure and thereby defer the delivery if it tries to use (for example) a
450 cached ldap connection that the parent has called unbind on.
452 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
453 of $address_data from the recipient verification was clobbered by the
456 13. The value of address_data from a sender verification is now available in
457 $sender_address_data in subsequent conditions in the ACL statement.
459 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
461 15. Added a new option "connect=<time>" to callout options, to set a different
464 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
465 was its contents. (It was OK if the option was not defined at all.)
467 17. A "Completed" log line is now written for messages that are removed from
468 the spool by the -Mrm option.
470 18. New variables $sender_verify_failure and $recipient_verify_failure contain
471 information about exactly what failed.
473 19. Added -dd to debug only the daemon process.
475 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
476 handles timeouts, both on the server side and network timeouts. Renamed the
477 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
479 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
480 It is now set to "smtps".
482 22. $host_address is now set to the target address during the checking of
485 23. When checking ignore_target_hosts for an ipliteral router, no host name was
486 being passed; this would have caused $sender_host_name to have been used if
487 matching the list had actually called for a host name (not very likely,
488 since this list is usually IP addresses). A host name is now passed as
491 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
492 code that specifies a non-restarting handler (typically sigaction() in
493 modern systems) in an attempt to fix a rare and obscure crash bug.
495 25. Narrowed the window for a race in the daemon that could cause it to ignore
496 SIGCHLD signals. This is not a major problem, because they are used only to
497 wake it up if nothing else does.
499 26. A malformed maildirsize file could cause Exim to calculate negative values
500 for the mailbox size or file count. Odd effects could occur as a result.
501 The maildirsize information is now recalculated if the size or filecount
504 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
505 support for a long time. Removed HAVE_SYS_VFS_H.
507 28. Installed the latest version of exipick from John Jetmore.
509 29. In an address list, if the pattern was not a regular expression, an empty
510 subject address (from a bounce message) matched only if the pattern was an
511 empty string. Non-empty patterns were not even tested. This was the wrong
512 because it is perfectly reasonable to use an empty address as part of a
513 database query. An empty address is now tested by patterns that are
514 lookups. However, all the other forms of pattern expect the subject to
515 contain a local part and a domain, and therefore, for them, an empty
516 address still always fails if the pattern is not itself empty.
518 30. Exim went into a mad DNS loop when attempting to do a callout where the
519 host was specified on an smtp transport, and looking it up yielded more
522 31. Re-factored the code for checking spool and log partition space into a
523 function that finds that data and another that does the check. The former
524 is then used to implement four new variables: $spool_space, $log_space,
525 $spool_inodes, and $log_inodes.
527 32. The RFC2047 encoding function was originally intended for short strings
528 such as real names; it was not keeping to the 75-character limit for
529 encoded words that the RFC imposes. It now respects the limit, and
530 generates multiple encoded words if necessary. To be on the safe side, I
531 have increased the buffer size for the ${rfc2047: expansion operator from
534 33. It is now permitted to omit both strings after an "if" condition; if the
535 condition is true, the result is "true". As before, when the second string
536 is omitted, a false condition yields an empty string. This makes it less
537 cumbersome to write custom ACL and router conditions.
539 34. Failure to deliver a bounce message always caused it to be frozen, even if
540 there was an errors_to setting on the router. The errors_to setting is now
543 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
544 canonical form (fully expanded) before being placed in
545 $sender_host_address.
547 36. The table in the code that translates DNS record types into text (T_A to
548 "A" for instance) was missing entries for NS and CNAME. It is just possible
549 that this could have caused confusion if both these types were looked up
550 for the same domain, because the text type is used as part of Exim's
551 per-process caching. But the chance of anyone hitting this buglet seems
554 37. The dnsdb lookup has been extended in a number of ways.
556 (1) There is a new type, "zns", which walks up the domain tree until it
557 finds some nameserver records. It should be used with care.
559 (2) There is a new type, "mxh", which is like "mx" except that it returns
560 just the host names, not the priorities.
562 (3) It is now possible to give a list of domains (or IP addresses) to be
563 looked up. The behaviour when one of the lookups defers can be
564 controlled by a keyword.
566 (4) It is now possible to specify the separator character for use when
567 multiple records are returned.
569 38. The dnslists ACL condition has been extended: it is now possible to supply
570 a list of IP addresses and/or domains to be looked up in a particular DNS
573 39. Added log_selector=+queue_time_overall.
575 40. When running the queue in the test harness, wait just a tad after forking a
576 delivery process, to get repeatability of debugging output.
578 41. Include certificate and key file names in error message when GnuTLS fails
579 to set them up, because the GnuTLS error message doesn't include the name
580 of the failing file when there is a problem reading it.
582 42. Allow both -bf and -bF in the same test run.
584 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
586 44. The "Exiscan patch" is now merged into the mainline Exim source.
588 45. Sometimes the final signoff response after QUIT could fail to get
589 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
590 < 0 before doing a fflush(). This bug looks as though it goes back to the
591 introduction of TLS in release 3.20, but "sometimes" must have been rare
592 because the tests only now provoked it.
594 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
595 (this can affect the format of dates).
597 47. exim_tidydb, when checking for the continued existence of a message for
598 which it has found a message-specific retry record, was not finding
599 messages that were in split spool directories. Consequently, it was
600 deleting retry records that should have stayed in existence.
602 48. Steve fixed some bugs in eximstats.
604 49. The SPA authentication driver was not abandoning authentication and moving
605 on to the next authenticator when an expansion was forced to fail,
606 contradicting the general specification for all authenticators. Instead it
607 was generating a temporary error. It now behaves as specified.
609 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
610 (the order specifies the preference for clients). The order is now AES256,
611 AES128, 3DES, ARCFOUR128.
613 51. Small patch to Sieve code - explicitly set From: when generating an
616 52. Exim crashed if a remote delivery caused a very long error message to be
617 recorded - for instance if somebody sent an entire SpamAssassin report back
618 as a large number of 550 error lines. This bug was coincidentally fixed by
619 increasing the size of one of Exim's internal buffers (big_buffer) that
620 happened as part of the Exiscan merge. However, to be on the safe side, I
621 have made the code more robust (and fixed the comments that describe what
624 53. Now that there can be additional text after "Completed" in log lines (if
625 the queue_time_overall log selector is set), a one-byte patch to exigrep
626 was needed to allow it to recognize "Completed" as not the last thing in
629 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
630 patch that reportedly fixes this has been added. I am not expert enough to
631 create a test for it. This is what the patch creator wrote:
633 "I found a little strange behaviour of ldap code when working with
634 Windows 2003 AD Domain, where users was placed in more than one
635 Organization Units. When I tried to give exim partial DN, the exit code
636 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
637 But simultaneously result of request was absolutely normal ldap result,
638 so I produce this patch..."
640 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
641 REFERENCE, so I have modified the code to exclude the patch when that macro
644 55. Some experimental protocols are using DNS PTR records for new purposes. The
645 keys for these records are domain names, not reversed IP addresses. The
646 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
647 leaves it alone. Component reversal etc. now happens only for IP addresses.
650 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
652 57. Double the size of the debug message buffer (to 2048) so that more of very
653 long debug lines gets shown.
655 58. The exicyclog utility now does better if the number of log files to keep
656 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
658 59. Two changes related to the smtp_active_hostname option:
660 (1) $smtp_active_hostname is now available as a variable.
661 (2) The default for smtp_banner uses $smtp_active_hostname instead
662 of $primary_hostname.
664 60. The host_aton() function is supposed to be passed a string that is known
665 to be a valid IP address. However, in the case of IPv6 addresses, it was
666 not checking this. This is a hostage to fortune. Exim now panics and dies
667 if the condition is not met. A case was found where this could be provoked
668 from a dnsdb PTR lookup with an IPv6 address that had more than 8
669 components; fortuitously, this particular loophole had already been fixed
670 by change 4.50/55 above.
672 If there are any other similar loopholes, the new check in host_aton()
673 itself should stop them being exploited. The report I received stated that
674 data on the command line could provoke the exploit when Exim was running as
675 exim, but did not say which command line option was involved. All I could
676 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
680 61. There was a buffer overflow vulnerability in the SPA authentication code
681 (which came originally from the Samba project). I have added a test to the
682 spa_base64_to_bits() function which I hope fixes it.
685 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
686 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
688 63. The daemon start-up calls getloadavg() while still root for those OS that
689 need the first call to be done as root, but it missed one case: when
690 deliver_queue_load_max is set with deliver_drop_privilege. This is
691 necessary for the benefit of the queue runner, because there is no re-exec
692 when deliver_drop_privilege is set.
694 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
697 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
698 was in use, was not putting the data itself into the right store pool;
699 consequently, it could be overwritten for a subsequent message in the same
700 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
703 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
705 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
706 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
707 the function were treating the return as a boolean value, which happened to
708 work because 0=false and not-0=true, but is not correct code.
710 68. The host_aton() function was not handling scoped IPv6 addresses (those
711 with, for example, "%eth0" on the end) correctly.
713 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
714 negated items (that is, ~something) in unsigned ints. Some compilers
715 apparently mutter when there is no cast.
717 70. If an address verification called from an ACL failed, and did not produce a
718 user-specific message (i.e. there was only a "system" message), nothing was
719 put in $acl_verify_message. In this situation, it now puts the system
722 71. Change 4.23/11 added synchronization checking at the start of an SMTP
723 session; change 4.31/43 added the unwanted input to the log line - except
724 that it did not do this in the start of session case. It now does.
726 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
727 This is wrong and can cause the other end to generate a synchronization
728 error if it is another Exim or anything else that does the synchronization
729 check. A QUIT command is no longer sent after a timeout.
731 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
734 74. The defer_ok option of callout verification was not working if it was used
735 when verifying addresses in header lines, that is, for this case:
737 verify = header_sender/callout=defer_ok
739 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
740 those file descriptors could be used for SMTP connections. If anything
741 wrote to stderr (the example that came up was "warn" in embedded Perl), it
742 could be sent to the SMTP client, causing chaos. The daemon now opens
743 stdin, stdout, and stderr to /dev/null when it puts itself into the
746 76. Arrange for output from Perl's "warn" command to be written to Exim's main
747 log by default. The user can override this with suitable Perl magic.
749 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
750 the log message when discard triggers, was not working for the DATA ACL or
751 for the non-SMTP ACL.
753 78. Error message wording change in sieve.c.
755 79. If smtp_accept_max_per_host was set, the number of connections could be
756 restricted to fewer than expected, because the daemon was trying to set up
757 a new connection before checking whether the processes handling previous
758 connections had finished. The check for completed processes is now done
759 earlier. On busy systems, this bug wouldn't be noticed because something
760 else would have woken the daemon, and it would have reaped the completed
763 80. If a message was submitted locally by a user whose login name contained one
764 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
765 It caused a spool format error. I have fixed the spool reading code. A
766 related problem was that the "from" clause in the Received: line became
767 illegal because of the space(s). It is now covered by ${quote_local_part.
769 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
772 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
774 83. Patch from Sieve maintainer to fix unterminated string problem in
777 84. Some minor changes to the Linux configuration files to help with other
778 OS variants using glibc.
780 85. One more patch for Sieve to update vacation handling to latest spec.
783 ----------------------------------------------------
784 See the note above about the 4.44 and 4.50 releases.
785 ----------------------------------------------------
791 1. Change 4.43/35 introduced a bug that caused file counts to be
792 incorrectly computed when quota_filecount was set in an appendfile
795 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
796 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
799 3. Give more explanation in the error message when the command for a transport
800 filter fails to execute.
802 4. There are several places where Exim runs a non-Exim command in a
803 subprocess. The SIGUSR1 signal should be disabled for these processes. This
804 was being done only for the command run by the queryprogram router. It is
805 now done for all such subprocesses. The other cases are: ${run, transport
806 filters, and the commands run by the lmtp and pipe transports.
808 5. Some older OS have a limit of 256 on the maximum number of file
809 descriptors. Exim was using setrlimit() to set 1000 as a large value
810 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
811 systems. I've change it so that if it can't get 1000, it tries for 256.
813 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
814 was an oversight, and furthermore, ever since the addition of extra
815 controls (e.g. 4.43/32), the checks on when to allow different forms of
816 "control" were broken. There should now be diagnostics for all cases when a
817 control that does not make sense is encountered.
819 7. $recipients is now available in the predata ACL (oversight).
821 8. Tidy the search cache before the fork to do a delivery from a message
822 received from the command line. Otherwise the child will trigger a lookup
823 failure and thereby defer the delivery if it tries to use (for example) a
824 cached ldap connection that the parent has called unbind on.
826 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
827 of $address_data from the recipient verification was clobbered by the
830 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
831 was its contents. (It was OK if the option was not defined at all.)
833 11. A "Completed" log line is now written for messages that are removed from
834 the spool by the -Mrm option.
836 12. $host_address is now set to the target address during the checking of
839 13. When checking ignore_target_hosts for an ipliteral router, no host name was
840 being passed; this would have caused $sender_host_name to have been used if
841 matching the list had actually called for a host name (not very likely,
842 since this list is usually IP addresses). A host name is now passed as
845 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
846 code that specifies a non-restarting handler (typically sigaction() in
847 modern systems) in an attempt to fix a rare and obscure crash bug.
849 15. Narrowed the window for a race in the daemon that could cause it to ignore
850 SIGCHLD signals. This is not a major problem, because they are used only to
851 wake it up if nothing else does.
853 16. A malformed maildirsize file could cause Exim to calculate negative values
854 for the mailbox size or file count. Odd effects could occur as a result.
855 The maildirsize information is now recalculated if the size or filecount
858 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
859 support for a long time. Removed HAVE_SYS_VFS_H.
861 18. Updated exipick to current release from John Jetmore.
863 19. Allow an empty sender to be matched against a lookup in an address list.
864 Previously the only cases considered were a regular expression, or an
867 20. Exim went into a mad DNS lookup loop when doing a callout where the
868 host was specified on the transport, if the DNS lookup yielded more than
871 21. The RFC2047 encoding function was originally intended for short strings
872 such as real names; it was not keeping to the 75-character limit for
873 encoded words that the RFC imposes. It now respects the limit, and
874 generates multiple encoded words if necessary. To be on the safe side, I
875 have increased the buffer size for the ${rfc2047: expansion operator from
878 22. Failure to deliver a bounce message always caused it to be frozen, even if
879 there was an errors_to setting on the router. The errors_to setting is now
882 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
883 canonical form (fully expanded) before being placed in
884 $sender_host_address.
886 24. Updated eximstats to version 1.33
888 25. Include certificate and key file names in error message when GnuTLS fails
889 to set them up, because the GnuTLS error message doesn't include the name
890 of the failing file when there is a problem reading it.
892 26. Expand error message when OpenSSL has problems setting up cert/key files.
895 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
896 (this can affect the format of dates).
898 28. exim_tidydb, when checking for the continued existence of a message for
899 which it has found a message-specific retry record, was not finding
900 messages that were in split spool directories. Consequently, it was
901 deleting retry records that should have stayed in existence.
903 29. eximstats updated to version 1.35
904 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
905 1.35 - bugfix such that pie charts by volume are generated correctly
907 30. The SPA authentication driver was not abandoning authentication and moving
908 on to the next authenticator when an expansion was forced to fail,
909 contradicting the general specification for all authenticators. Instead it
910 was generating a temporary error. It now behaves as specified.
912 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
913 (the order specifies the preference for clients). The order is now AES256,
914 AES128, 3DES, ARCFOUR128.
916 31. Small patch to Sieve code - explicitly set From: when generating an
919 32. Exim crashed if a remote delivery caused a very long error message to be
920 recorded - for instance if somebody sent an entire SpamAssassin report back
921 as a large number of 550 error lines. This bug was coincidentally fixed by
922 increasing the size of one of Exim's internal buffers (big_buffer) that
923 happened as part of the Exiscan merge. However, to be on the safe side, I
924 have made the code more robust (and fixed the comments that describe what
927 33. Some experimental protocols are using DNS PTR records for new purposes. The
928 keys for these records are domain names, not reversed IP addresses. The
929 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
930 leaves it alone. Component reversal etc. now happens only for IP addresses.
933 34. The host_aton() function is supposed to be passed a string that is known
934 to be a valid IP address. However, in the case of IPv6 addresses, it was
935 not checking this. This is a hostage to fortune. Exim now panics and dies
936 if the condition is not met. A case was found where this could be provoked
937 from a dnsdb PTR lookup with an IPv6 address that had more than 8
938 components; fortuitously, this particular loophole had already been fixed
939 by change 4.50/55 or 4.44/33 above.
941 If there are any other similar loopholes, the new check in host_aton()
942 itself should stop them being exploited. The report I received stated that
943 data on the command line could provoke the exploit when Exim was running as
944 exim, but did not say which command line option was involved. All I could
945 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
949 35. There was a buffer overflow vulnerability in the SPA authentication code
950 (which came originally from the Samba project). I have added a test to the
951 spa_base64_to_bits() function which I hope fixes it.
954 36. The daemon start-up calls getloadavg() while still root for those OS that
955 need the first call to be done as root, but it missed one case: when
956 deliver_queue_load_max is set with deliver_drop_privilege. This is
957 necessary for the benefit of the queue runner, because there is no re-exec
958 when deliver_drop_privilege is set.
960 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
961 was in use, was not putting the data itself into the right store pool;
962 consequently, it could be overwritten for a subsequent message in the same
963 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
966 38. Sometimes the final signoff response after QUIT could fail to get
967 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
968 < 0 before doing a fflush(). This bug looks as though it goes back to the
969 introduction of TLS in release 3.20, but "sometimes" must have been rare
970 because the tests only now provoked it.
976 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
977 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
978 Now it returns an integer. A number of calls were still expecting a T/F
979 return. Fortuitously, in all cases, the tests worked in OK situations,
980 which is the norm. However, things would have gone wrong on any write
981 failures on the smtp file descriptor. This function is used when sending
982 messages over SMTP and also when doing verify callouts.
984 2. When Exim is called to do synchronous delivery of a locally submitted
985 message (the -odf or -odi options), it no longer closes stderr before doing
988 3. Implemented the mua_wrapper option.
990 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
992 5. Implemented the functions header_remove(), header_testname(),
993 header_add_at_position(), and receive_remove_recipient(), and exported them
996 6. If an ACL "warn" statement specified the addition of headers, Exim already
997 inserted X-ACL-Warn: at the start if there was no header name. However, it
998 was not making this test for the second and subsequent header lines if
999 there were newlines in the string. This meant that an invalid header could
1000 be inserted if Exim was badly configured.
1002 7. Allow an ACL "warn" statement to add header lines at the start or after all
1003 the Received: headers, as well as at the end.
1005 8. Added the rcpt_4xx retry error code.
1007 9. Added postmaster_mailfrom=xxx to callout verification option.
1009 10. Added mailfrom=xxxx to the callout verification option, for verify=
1012 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1013 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1015 12. Inserted some casts to stop certain compilers warning when using pointer
1016 differences as field lengths or precisions in printf-type calls (mostly
1017 affecting debugging statements).
1019 13. Added optional readline() support for -be (dynamically loaded).
1021 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1022 same clock tick as a message's arrival, so that its received time was the
1023 same as the "first fail" time on the retry record, and that message
1024 remained on the queue past the ultimate address timeout, every queue runner
1025 would try a delivery (because it was past the ultimate address timeout) but
1026 after another failure, the ultimate address timeout, which should have then
1027 bounced the address, did not kick in. This was a "< instead of <=" error;
1028 in most cases the first failure would have been in the next clock tick
1029 after the received time, and all would be well.
1031 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1032 being recognized when the domain list was tested by the match_domain
1033 condition in an expansion string.
1035 16. Added the ${str2b64: operator.
1037 17. Exim was always calling setrlimit() to set a large limit for the number of
1038 processes, without checking whether the existing limit was already
1039 adequate. (It did check for the limit on file descriptors.) Furthermore,
1040 errors from getrlimit() and setrlimit() were being ignored. Now they are
1041 logged to the main and panic logs, but Exim does carry on, to try to do its
1042 job under whatever limits there are.
1044 18. Imported PCRE 5.0.
1046 19. Trivial typo in log message " temporarily refused connection" (the leading
1049 20. If the log selector return_path_on_delivery was set and an address was
1050 redirected to /dev/null, the delivery process crashed because it assumed
1051 that a return path would always be set for a "successful" delivery. In this
1052 case, the whole delivery is bypassed as an optimization, and therefore no
1055 21. Internal re-arrangement: the function for sending a challenge and reading
1056 a response while authentication was assuming a zero-terminated challenge
1057 string. It's now changed to take a pointer and a length, to allow for
1058 binary data in such strings.
1060 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1062 23. Exim was not respecting finduser_retries when seeking the login of the
1063 uid under which it was called; it was always trying 10 times. (The default
1064 setting of finduser_retries is zero.) Also, it was sleeping after the final
1065 failure, which is pointless.
1067 24. Implemented tls_on_connect_ports.
1069 25. Implemented acl_smtp_predata.
1071 26. If the domain in control=submission is set empty, Exim assumes that the
1072 authenticated id is a complete email address when it generates From: or
1073 Sender: header lines.
1075 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1076 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1077 chown and chgrp in /bin and hostname in /usr/bin.
1079 28. Exim was keeping the "process log" file open after each use, just as it
1080 does for the main log. This opens the possibility of it remaining open for
1081 long periods when the USR1 signal hits a daemon. Occasional processlog
1082 errors were reported, that could have been caused by this. Anyway, it seems
1083 much more sensible not to leave this file open at all, so that is what now
1086 29. The long-running daemon process does not normally write to the log once it
1087 has entered its main loop, and it closes the log before doing so. This is
1088 so that log files can straightforwardly be renamed and moved. However,
1089 there are a couple of unusual error situations where the daemon does write
1090 log entries, and I had neglected to close the log afterwards.
1092 30. The text of an SMTP error response that was received during a remote
1093 delivery was being truncated at 512 bytes. This is too short for some of
1094 the long messages that one sometimes sees. I've increased the limit to
1097 31. It is now possible to make retry rules that apply only when a message has a
1098 specific sender, in particular, an empty sender.
1100 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1101 it possible to be selective about when SMTP synchronization is enforced.
1103 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1105 32. Implemented hosts_connection_nolog.
1107 33. Added an ACL for QUIT.
1109 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1112 35. Added mailbox_size and mailbox_filecount to appendfile.
1114 36. Added control = no_multiline_responses to ACLs.
1116 37. There was a bug in the logic of the code that waits for the clock to tick
1117 in the case where the clock went backwards by a substantial amount such
1118 that the microsecond fraction of "now" was more than the microsecond
1119 fraction of "then" (but the whole seconds number was less).
1121 38. Added support for the libradius Radius client library this is found on
1122 FreeBSD (previously only the radiusclient library was supported).
1128 1. When certain lookups returned multiple values in the form name=value, the
1129 quoting of the values was not always being done properly. Specifically:
1130 (a) If the value started with a double quote, but contained no whitespace,
1132 (b) If the value contained whitespace other than a space character (i.e.
1133 tabs or newlines or carriage returns) it was not quoted.
1134 This fix has been applied to the mysql and pgsql lookups by writing a
1135 separate quoting function and calling it from the lookup code. The fix
1136 should probably also be applied to nisplus, ibase and oracle lookups, but
1137 since I cannot test any of those, I have not disturbed their existing code.
1139 2. A hit in the callout cache for a specific address caused a log line with no
1140 reason for rejecting RCPT. Now it says "Previous (cached) callout
1141 verification failure".
1143 3. There was an off-by-one bug in the queryprogram router. An over-long
1144 return line was truncated at 256 instead of 255 characters, thereby
1145 overflowing its buffer with the terminating zero. As well as fixing this, I
1146 have increased the buffer size to 1024 (and made a note to document this).
1148 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1149 when Exim is waiting for an SMTP response from a remote server, Exim
1150 restarts its select() call on the socket, thereby resetting its timeout.
1151 This is not a problem when such interrupts are rare. Somebody set up a cron
1152 job to run exiwhat every 2 minutes, which is less than the normal select()
1153 timeout (5 or 10 minutes). This meant that the select() timeout never
1154 kicked in because it was always reset. I have fixed this by comparing the
1155 time when an interrupt arrives with the time at the start of the first call
1156 to select(). If more time than the timeout has elapsed, the interrupt is
1157 treated as a timeout.
1159 5. Some internal re-factoring in preparation for the addition of Sieve
1160 extensions (by MH). In particular, the "personal" test is moved to a
1161 separate function, and given an option for scanning Cc: and Bcc: (which is
1162 not set for Exim filters).
1164 6. When Exim created an email address using the login of the caller as the
1165 local part (e.g. when creating a From: or Sender: header line), it was not
1166 quoting the local part when it contained special characters such as @.
1168 7. Installed new OpenBSD configuration files.
1170 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1171 try to make them clearer.
1173 9. Callout options, other than the timeout value, were being ignored when
1174 verifying sender addresses in header lines. For example, when using
1176 verify = header_sender/callout=no_cache
1178 the cache was (incorrectly) being used.
1180 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1181 only the Cygwin environment.
1183 11. When return_path_on_delivery was set as a log selector, if different remote
1184 addresses in the same message used different return paths and parallel
1185 remote delivery occurred, the wrong values would sometimes be logged.
1186 (Whenever a remote delivery process finished, the return path value from
1187 the most recently started remote delivery process was logged.)
1189 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1190 lines when AUTH and/or TLS are in use. This is the "received protocol"
1191 field. Exim used to use "asmtp" for authenticated SMTP, without any
1192 indication (in the protocol name) for TLS use. Now it follows the RFC and
1193 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1194 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1195 names appear in log lines as well as in Received: header lines.
1197 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1198 extensions, and comparison tests, and to fix some bugs.
1200 14. Changes to the "personal" filter test:
1202 (1) The test was buggy in that it was just doing the equivalent of
1203 "contains" tests on header lines. For example, if a user's address was
1204 anne@some.where, the "personal" test would incorrectly be true for
1206 To: susanne@some.where
1208 This test is now done by extracting each address from the header in turn,
1209 and checking the entire address. Other tests that are part of "personal"
1210 are now done using regular expressions (for example, to check local parts
1211 of addresses in From: header lines).
1213 (2) The list of non-personal local parts in From: addresses has been
1214 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1215 taken from the Sieve specification recommendations.
1217 (3) If the message contains any header line starting with "List-" it is
1218 treated as non-personal.
1220 (4) The test for "circular" in the Subject: header line has been removed
1221 because it now seems ill-conceived.
1223 15. Minor typos in src/EDITME comments corrected.
1225 16. Installed latest exipick from John Jetmore.
1227 17. If headers_add on a router specified a text string that was too long for
1228 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1229 of string_sprintf() is now avoided.
1231 18. $message_body_size was not set (it was always zero) when running the DATA
1232 ACL and the local_scan() function.
1234 19. For the "mail" command in an Exim filter, no default was being set for
1235 the once_repeat time, causing a random time value to be used if "once" was
1236 specified. (If the value happened to be <= 0, no repeat happened.) The
1237 default is now 0s, meaning "never repeat". The "vacation" command was OK
1238 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1239 (I found it when inspecting the code).
1241 20. There is now an overall timeout for performing a callout verification. It
1242 defaults to 4 times the callout timeout, which applies to individual SMTP
1243 commands during the callout. The overall timeout applies when there is more
1244 than one host that can be tried. The timeout is checked before trying the
1245 next host. This prevents very long delays if there are a large number of
1246 hosts and all are timing out (e.g. when the network connections are timing
1247 out). The value of the overall timeout can be changed by specifying an
1248 additional sub-option for "callout", called "maxwait". For example:
1250 verify = sender/callout=5s,maxwait=20s
1252 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1253 to the end before writing, but this should make it even safer).
1255 22. Exim was forgetting that it had advertised PIPELINING for the second and
1256 subsequent messages on an SMTP connection. It was also not resetting its
1257 memory on STARTTLS and an internal HELO.
1259 23. When Exim logs an SMTP synchronization error within a session, it now
1260 records whether PIPELINING has been advertised or not.
1262 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1263 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1264 rather than long int.
1266 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1268 26. Added the never_mail option to autoreply.
1274 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1275 crash if the getsockname() call failed; this can happen if a connection is
1276 closed very soon after it is established. The problem was simply in the
1277 order in which certain operations were done, causing Exim to try to write
1278 to the SMTP stream before it had set up the file descriptor. The bug has
1279 been fixed by making things happen in the correct order.
1285 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1286 before the connection was closed, thus losing the rejection response.
1288 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1289 some early Solaris releases, but causes trouble in current releases where
1290 socklen_t is defined.
1292 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1295 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1298 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1299 that releasing the top of it at the end releases what was used for sub-
1300 expansions (unless the block got too big). However, discard this block if
1301 the first thing is a variable or header, so that we can use its block when
1302 it is dynamic (useful for very large $message_headers, for example).
1304 6. Lookups now cache *every* query, not just the most recent. A new, separate
1305 store pool is used for this. It can be recovered when all lookup caches are
1306 flushed. Lookups now release memory at the end of their result strings.
1307 This has involved some general refactoring of the lookup sources.
1309 7. Some code has been added to the store_xxx() functions to reduce the amount
1310 of flapping under certain conditions.
1312 8. log_incoming_interface used to affect only the <= reception log lines. Now
1313 it causes the local interface and port to be added to several more SMTP log
1314 lines, for example "SMTP connection from", and rejection lines.
1316 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1318 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1320 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1321 could be overwritten at the end of the current message (or the start of a
1322 new message if it was set in a HELO ACL). The value is now preserved for
1323 the duration of the SMTP connection.
1325 12. If a transport had a headers_rewrite setting, and a matching header line
1326 contained an unqualified address, that address was qualified, even if it
1327 did not match any rewriting rules. The underlying bug was that the values
1328 of the flags that permit the existence of unqualified sender and recipient
1329 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1330 non-local messages, and by -bnq for local messages) were not being
1331 preserved with the message after it was received.
1333 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1334 "next input=" as part of the text comprising the host identity instead of
1335 the correct text. The code was using the same buffer for two different
1336 strings. However, depending on which order the printing function evaluated
1337 its arguments, the bug did not always show up. Under Linux, for example, my
1338 test suite worked just fine.
1340 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1341 doesn't work with some older versions of Perl. It has been changed to "my",
1342 which in any case is probably the better facility to use.
1344 15. A really picky compiler found some instances of statements for creating
1345 error messages that either had too many or two few arguments for the format
1348 16. The size of the buffer for calls to the DNS resolver has been increased
1349 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1350 for addresses that have a lot of PTR records. This alleviates a problem; it
1351 does not fully solve it.
1353 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1354 buffer now truncates the list and logs the incident, which is the same
1355 action as happens when Exim is looking up a host name and its aliases.
1356 Previously in this situation something unpredictable would happen;
1357 sometimes it was "internal error: store_reset failed".
1359 18. If a server dropped the connection unexpectedly when an Exim client was
1360 using GnuTLS and trying to read a response, the client delivery process
1361 crashed while trying to generate an error log message.
1363 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1364 string, for example:
1366 warn message = H1: something\nH2: something
1368 the text was added as a single header line from Exim's point of view
1369 though it ended up OK in the delivered message. However, searching for the
1370 second and subsequent header lines using $h_h2: did not work. This has been
1371 fixed. Similarly, if a system filter added multiple headers in this way,
1372 the routers could not see them.
1374 20. Expanded the error message when iplsearch is called with an invalid key to
1375 suggest using net-iplsearch in a host list.
1377 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1378 ACLs are no longer actually imposed (and a message to that effect is
1381 22. If a "gecos" field in a passwd entry contained escaped characters, in
1382 particular, if it contained a \" sequence, Exim got it wrong when building
1383 a From: or a Sender: header from that name. A second bug also caused
1384 incorrect handling when an unquoted " was present following a character
1385 that needed quoting.
1387 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1388 was not being matched caselessly.
1390 24. Arranged for all hyphens in the exim.8 source to be escaped with
1393 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1394 properly. Recipient callout cache records were still being keyed to include
1395 the sender, even when use_sender was set false. This led to far more
1396 callouts that were necessary. The sender is no longer included in the key
1397 when use_sender is false.
1399 26. Added "control = submission" modifier to ACLs.
1401 27. Added the ${base62d: operator to decode base 62 numbers.
1403 28. dnsdb lookups can now access SRV records.
1405 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1406 the configuration file.
1408 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1409 (-v) mode. This makes the output for a verbose queue run more intelligible.
1411 31. Added a use_postmaster feature to recipient callouts.
1413 32. Added the $body_zerocount variable, containing the number of binary zero
1414 bytes in the message body.
1416 33. The time of last modification of the "new" subdirectory is now used as the
1417 "mailbox time last read" when there is a quota error for a maildir
1420 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1422 35. Added +ignore_unknown as a special item in host lists.
1424 36. Code for decoding IPv6 addresses in host lists is now included, even if
1425 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1426 address was recognized as an IP address, but was then not correctly decoded
1427 into binary, causing unexpected and incorrect effects when compared with
1434 1. Very minor rewording of debugging text in manualroute to say "list of
1435 hosts" instead of "hostlist".
1437 2. If verify=header_syntax was set, and a header line with an unqualified
1438 address (no domain) and a large number of spaces between the end of the
1439 name and the colon was received, the reception process suffered a buffer
1440 overflow, and (when I tested it) crashed. This was caused by some obsolete
1441 code that should have been removed. The fix is to remove it!
1443 3. When running in the test harness, delay a bit after writing a bounce
1444 message to get a bit more predictability in the log output.
1446 4. Added a call to search_tidyup() just before forking a reception process. In
1447 theory, someone could use a lookup in the expansion of smtp_accept_max_
1448 per_host which, without the tidyup, could leave open a database connection.
1450 5. Added the variables $recipient_data and $sender_data which get set from a
1451 lookup success in an ACL "recipients" or "senders" condition, or a router
1452 "senders" option, similar to $domain_data and $local_part_data.
1454 6. Moved the writing of debug_print from before to after the "senders" test
1457 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1458 problems for message scanning, either using a data ACL, or using
1459 local_scan() because the Received: header was not generated till after they
1460 were called (in order to set the time as the time of reception completion).
1461 I have revised the way this works. The header is now generated after the
1462 body is received, but before the ACL or local_scan() are called. After they
1463 are run, the timestamp in the header is updated.
1469 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1470 before starting a queue runner without re-exec. This happened only when
1471 deliver_drop_privilege was set or when the Exim user was set to root. The
1472 effect of the bug was that timeouts during subsequent deliveries caused
1473 crashes instead of being properly handled. The handler is now left at its
1474 default (and expected) setting.
1476 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1477 message, again when deliver_drop_privilege is set or Exim is run as root.
1478 The bug described in (1) was not present in this case, but the tidying up
1479 of the other signals was missing. I have made the two cases consistent.
1481 3. The ignore_target_hosts setting on a manualroute router was being ignored
1482 for hosts that were looked up using the /MX notation.
1484 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1487 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1488 operated on the sender address. After changing the $sender_address to <>
1489 for the sender address verify, Exim was re-instated it as the original
1490 (before rewriting) address, but remembering that it had rewritten it, so it
1491 wasn't rewriting it again. This bug also had the effect of breaking the
1492 sender address verification caching when the sender address was rewritten.
1494 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1495 This has been changed so that if the ip literal address matches
1496 ignore_target_hosts, the router declines.
1498 7. Added expansion conditions match_domain, match_address, and match_local_
1499 part (NOT match_host).
1501 8. The placeholder for the Received: header didn't have a length field set.
1503 9. Added code to Exim itself and to exim_lock to test for a specific race
1504 condition that could lead to file corruption when using MBX delivery. The
1505 issue is with the lockfile that is created in /tmp. If this file is removed
1506 after a process has opened it but before that process has acquired a lock,
1507 there is the potential for a second process to recreate the file and also
1508 acquire a lock. This could lead to two Exim processes writing to the file
1509 at the same time. The added code performs the same test as UW imapd; it
1510 checks after acquiring the lock that its file descriptor still refers to
1511 the same named file.
1513 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1514 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1517 11. Added the smtp_active_hostname option. If used, this will typically be made
1518 to depend on the incoming interface address. Because $interface_address is
1519 not set up until the daemon has forked a reception process, error responses
1520 that can happen earlier (such as "too many connections") no longer contain
1523 12. If an expansion in a condition on a "warn" statement fails because a lookup
1524 defers, the "warn" statement is abandoned, and the next ACL statement is
1525 processed. Previously this caused the whole ACL to be aborted.
1527 13. Added the iplsearch lookup type.
1529 14. Added ident_timeout as a log selector.
1531 15. Added tls_certificate_verified as a log selector.
1533 16. Added a global option tls_require_ciphers (compare the smtp transport
1534 option of the same name). This controls incoming TLS connections.
1536 17. I finally figured out how to make tls_require_ciphers do a similar thing
1537 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1538 before starting the TLS session.
1540 18. Tabs are now shown as \t in -bP output.
1542 19. If the log selector return_path_on_delivery was set, Exim crashed when
1543 bouncing a message because it had too many Received: header lines.
1545 20. If two routers both had headers_remove settings, and the first one included
1546 a superfluous trailing colon, the final name in the first list and the
1547 first name in the second list were incorrectly joined into one item (with a
1548 colon in the middle).
1554 1. Added -C and -D options to the exinext utility, mainly to make it easier
1555 to include in the automated testing, but these could be helpful when
1556 multiple configurations are in use.
1558 2. The exinext utility was not formatting the output nicely when there was
1559 an alternate port involved in the retry record key, nor when there was a
1560 message id as well (for retries that were specific to a specific message
1561 and a specific host). It was also confused by IPv6 addresses, because of
1562 the additional colons they contain. I have fixed the IPv4 problem, and
1563 patched it up to do a reasonable job for IPv6.
1565 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1566 delivery, the log line now contains "pipelined" if PIPELINING was used.
1568 4. An SMTP transport process used to panic and die if the bind() call to set
1569 an explicit outgoing interface failed. This has been changed; it is now
1570 treated in the same way as a connect() failure.
1572 5. A reference to $sender_host_name in the part of a conditional expansion
1573 that was being skipped was still causing a DNS lookup. This no longer
1576 6. The def: expansion condition was not recognizing references to header lines
1577 that used bh_ and bheader_.
1579 7. Added the _cache feature to named lists.
1581 8. The code for checking quota_filecount in the appendfile transport was
1582 allowing one more file than it should have been.
1584 9. For compatibility with Sendmail, the command line option
1592 and sets the incoming protocol and host name (for trusted callers). The
1593 host name and its colon can be omitted when only the protocol is to be set.
1594 Note the Exim already has two private options, -pd and -ps, that refer to
1595 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1596 "s", but I don't think that's a major issue.
1598 10. A number of refactoring changes to the code, none of which should affect
1601 (a) The number of logging options was getting close to filling up the
1602 32-bit word that was used as a bit map. I have split them into two classes:
1603 those that are passed in the argument to log_write(), and those that are
1604 only ever tested independently outside of that function. These are now in
1605 separate 32-bit words, so there is plenty of room for expansion again.
1606 There is no change in the user interface or the logging behaviour.
1608 (b) When building, for example, log lines, the code previously used a
1609 macro that called string_cat() twice, in order to add two strings. This is
1610 not really sufficiently general. Furthermore, there was one instance where
1611 it was actually wrong because one of the argument was used twice, and in
1612 one call a function was used. (As it happened, calling the function twice
1613 did not affect the overall behaviour.) The macro has been replaced by a
1614 function that can join an arbitrary number of extra strings onto a growing
1617 (c) The code for expansion conditions now uses a table and a binary chop
1618 instead of a serial search (which was left over from when there were very
1619 few conditions). Also, it now recognizes conditions like "pam" even when
1620 the relevant support is not compiled in: a suitably worded error message is
1621 given if an attempt is made to use such a condition.
1623 11. Added ${time_interval:xxxxx}.
1625 12. A bug was causing one of the ddress fields not to be passed back correctly
1626 from remote delivery subprocesses. The field in question was not being
1627 subsequently used, so this caused to problems in practice.
1629 13. Added new log selectors queue_time and deliver_time.
1631 14. Might have fixed a bug in maildirsizefile handling that threw up
1632 "unexpected character" debug warnings, and recalculated the data
1633 unnecessarily. In any case, I expanded the warning message to give more
1636 15. Added the message "Restricted characters in address" to the statements in
1637 the default ACL that block characters like @ and % in local parts.
1639 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1640 Three changes have been made:
1642 (a) There was a serious bug; a negative response to MAIL caused the whole
1643 recipient domain to be cached as invalid, thereby blocking all messages
1644 to all local parts at the same domain, from all senders. This bug has
1645 been fixed. The domain is no longer cached after a negative response to
1646 MAIL if the sender used is not empty.
1648 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1651 (c) A new callout option, "use_sender" has been added for people who want
1652 the modified behaviour.
1658 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1661 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1662 indeed breaks things for older releases.
1664 3. Added additional logging to the case where there is a problem reading data
1665 from a filter that is running in a subprocess using a pipe, in order to
1666 try to track down a specific problem.
1668 4. Testing facility fudge: when running in the test harness and attempting
1669 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1670 getting "No route to host". Convert this to a timeout.
1672 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1675 6. Some OS don't have socklen_t but use size_t instead. This affects the
1676 fifth argument of getsockopt() amongst other things. This is now
1677 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1678 can be set for individual OS. I have set it for SunOS5, OSF1, and
1679 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1680 some earlier ones do not.
1682 7. Change 4.30/15 was not doing the test caselessly.
1684 8. The standard form for an IPv6 address literal was being rejected by address
1685 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1686 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1687 this, as well as the form without the "IPv6" on the front (but only when
1688 address literals are enabled, of course).
1690 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1692 10. Exim crashed if a message with an empty sender address specified by -f
1693 encountered a router with an errors_to setting. This could be provoked only
1694 by a command such as
1698 where an empty string was supplied; "<>" did not hit this bug.
1700 11. Installed PCRE release 4.5.
1702 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1703 remained set. It is now erased.
1705 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1706 times from message ids (which are base 36 rather than the normal 62).
1708 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1709 were being counted as actual protocol errors, and logged if the log
1710 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1711 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1712 and DATA following a set of rejected RCPTs do not count as protocol errors.
1713 In other words, Exim assumes they were pipelined, though this may not
1714 actually be the case. Of course, in all cases the client gets an
1715 appropriate error code.
1717 15. If a lookup fails in an ACL condition, a message about the failure may
1718 be available; it is used if testing the ACL cannot continue, because most
1719 such messages specify what the cause of the deferral is. However, some
1720 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1721 that caused an old message to be retained and used if a later statement
1722 caused a defer, replacing the real cause of the deferral.
1724 16. If an IP address had so many PTR records that the DNS lookup buffer
1725 was not large enough to hold them, Exim could crash while trying to process
1726 the truncated data. It now detects and logs this case.
1728 17. Further to 4.21/58, another change has been made: if (and only if) the
1729 first line of a message (the first header line) ends with CRLF, a bare LF
1730 in a subsequent header line has a space inserted after it, so as not to
1731 terminate the header.
1733 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1734 unnecessarily, used atoi() instead of strtol(), and didn't check the
1735 termination when getting file sizes from file names by regex.
1737 19. Completely re-implemented the support for maildirsize files, in the light
1738 of a number of problems with the previous contributed implementation
1739 (4.30/29). In particular:
1741 . If the quota is zero, the maildirsize file is maintained, but no quota is
1744 . If the maildir directory does not exist, it is created before any attempt
1745 to write a maildirsize file.
1747 . The quota value in the file is just a cache; if the quota is changed in
1748 the transport, the new value overrides.
1750 . A regular expression is available for excluding directories from the
1753 20. The autoreply transport checks the characters in options that define the
1754 message's headers; it allows continued headers, but it was checking with
1755 isspace() after an embedded newline instead of explicitly looking for a
1758 21. If all the "regular" hosts to which an address was routed had passed their
1759 expiry times, and had not reached their retry times, the address was
1760 bounced, even if fallback hosts were defined. Now Exim should go on to try
1763 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1764 equivalent code in the SMTP transport. Some hosts send humungous responses
1765 to HELO/EHLO, more than 1024 it seems.
1767 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1768 gives compiler warnings in some environments. I've now done it "properly",
1771 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1772 (because of problems with the built-in one) was declared to return uschar *
1773 instead of char *, causing compiler failure.
1775 25. Fixed a file descriptor leak when processing alias/forward files.
1777 26. Fixed a minor format string issue in dbfn.c.
1779 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1781 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1782 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1785 29. When a sender address is verified, it is cached, to save repeating the test
1786 when there is more than one recipient in a message. However, when the
1787 verification involves a callout, it is possible for different callout
1788 options to be set for different recipients. It is too complicated to keep
1789 track of this in the cache, so now Exim always runs a verification when a
1790 callout is required, relying on the callout cache for the optimization.
1791 The overhead is duplication of the address routing, but this should not be
1794 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1795 to be verified with callout=postmaster, and the main callout worked but the
1796 postmaster check failed, the verification correctly failed. However, if a
1797 subsequent RCPT command asked for sender verification *without* the
1798 postmaster check, incorrect caching caused this verification also to fail,
1801 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1802 it was not caching the DNS options (qualify_single, search_parents) that
1803 were used when the lookup failed. A subsequent lookup with different
1804 options therefore always gave the same answer, though there were cases
1805 where it should not have. (Example: a "domains = !$mx_any" option on a
1806 dnslookup router: the "domains" option is always processed without any
1807 widening, but the router might have qualify_single set.) Now Exim uses the
1808 cached value only when the same options are set.
1810 32. Added John Jetmore's "exipick" utility to the distribution.
1812 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1813 than a timeout (e.g. a certificate is required, and is not provided), an
1814 Exim server now closes the connection immediately. Previously it waited for
1815 the client to close - but if the client is SSL, it seems that they each
1816 wait for each other, leading to a delay before one of them times out.
1818 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1819 maintained 0.8.x compatibility because I don't think many are using it, and
1820 it is clearly obsolete.
1822 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1825 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1826 client certificate was expired. A simple patch fixes this, though I don't
1827 understand the full logic of why the verify callback is called multiple
1830 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1831 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1832 which causes problems with some clients (such as the Certicom SSL Plus
1833 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1834 disables the coutermeasure allowing Eudora to connect."
1836 38. Exim was not checking that a write() to a log file succeeded. This could
1837 lead to Bad Things if a log got too big, in particular if it hit a file
1838 size limit. Exim now panics and dies if it cannot write to a log file, just
1839 as it does if it cannot open a log file.
1841 39. Modified OS/Makefile-Linux so that it now contains
1843 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1845 The two -D definitions ensure that Exim is compiled with large file
1846 support, which makes it possible to handle log files that are bigger than
1849 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1850 instance) a domain was checked against a named list that involved a lookup,
1851 causing $domain_data to be set, then another domain was checked against the
1852 same list, then the first domain was re-checked, the value of $domain_data
1853 after the final check could be wrong. In particular, if the second check
1854 failed, it could be set empty. This bug probably also applied to
1857 41. The strip_trailing_dot option was not being applied to the address given
1858 with the -f command-line option.
1860 42. The code for reading a message's header from the spool was incrementing
1861 $received_count, but never initializing it. This meant that the value was
1862 incorrect (doubled) while delivering a message in the same process in which
1863 it was received. In the most common configuration of Exim, this never
1864 happens - a fresh exec is done - but it can happen when
1865 deliver_drop_privilege is set.
1867 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1868 it now includes up to 150 characters of the unexpected data in the log
1871 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1872 and building data strings. The size of both of these buffers was 10 000
1873 bytes - far larger than anybody would *ever* want, thought I. Needless to
1874 say, somebody hit the limit. I have increased the maximum line length to
1875 20 000 and the maximum data length of concatenated lines to 100 000. I have
1876 also fixed two bugs, because there was no checking on these buffers. Tsk,
1877 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1878 buffer is too small.
1880 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1881 lsearch lookups. Now it does.
1883 46. When parsing a route_list item in a manualroute router, a fixed-length
1884 buffer was used for the list of hosts. I made this 1024 bytes long,
1885 thinking that nobody would ever have a list of hosts that long. Wrong.
1886 Somebody had a whole pile of complicated expansion conditions, and the
1887 string was silently truncated, leading to an expansion error. It turns out
1888 that it is easier to change to an unlimited length (owing to other changes
1889 that have happened since this code was originally written) than to build
1890 structure for giving a limitation error. The length of the item that
1891 expands into the list of hosts is now unlimited.
1893 47. The lsearch lookup could not handle data where the length of text line was
1894 more than 4095 characters. Such lines were truncated, leading to shortened
1895 data being returned. It should now handle lines of any length.
1897 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1898 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1901 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1902 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1903 no longer happens in comment lines. A list of replacements is now placed
1904 at the head of all of the source files, except those whose only change is
1905 to replace PERL_COMMAND in the very first #! line.
1907 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1908 messages on the queue, with a bottom-up merge sort, using code contributed
1909 by Michael Haardt. This should make operations like -bp somewhat faster on
1910 large queues. It won't affect queue runners, except when queue_run_in_order
1913 51. Installed eximstats 1.31 in the distribution.
1915 52. Added support for SRV lookups to the dnslookup router.
1917 53. If an ACL referred to $message_body or $message_body_end, the value was not
1918 reset for any messages that followed in the same SMTP session.
1920 54. The store-handling optimization for building very long strings was not
1921 differentiating between the different store pools. I don't think this
1922 actually made any difference in practice, but I've tidied it.
1924 55. While running the routers to verify a sender address, $sender_address
1925 was still set to the sender address. This is wrong, because when routing to
1926 send a bounce to the sender, it would be empty. Therefore, I have changed
1927 it so that, while verifying a sender address, $sender_address is set to <>.
1928 (There is no change to what happens when verifying a recipient address.)
1930 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1931 target A or AAAA records (if not already returned) without resetting the
1932 qualify_single or search_parents options of the DNS resolver. These are
1933 inappropriate in this case because the targets of MX and SRV records must
1934 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1935 target that, when qualified, matched something in the local domain. These
1936 two options are now turned off when doing these lookups.
1938 57. It seems that at least some releases of Reiserfs (which does not have the
1939 concept of a fixed number of inodes) returns zero and not -1 for the
1940 number of available inodes. This interacted badly with check_spool_inodes,
1941 which assumed that -1 was the "no such thing" setting. What I have done is
1942 to check that the total number of inodes is greater than zero before doing
1943 the test of how many are available.
1945 58. When a "warn" ACL statement has a log_message modifier, the message is
1946 remembered, and not repeated. This is to avoid a lot of repetition when a
1947 message has many recipients that cause the same warning to be written.
1948 Howewer, Exim was preserving the list of already written lines for an
1949 entire SMTP session, which doesn't seem right. The memory is now reset if a
1950 new message is started.
1952 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1953 debugging output unless log_rewrite was also set.
1955 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1956 of (int)(handle) when we know that handle contains (void *)(-1).
1958 61. The Exim daemon panic-logs an error return when it closes the incoming
1959 connection. However "connection reset by peer" seems to be common, and
1960 isn't really an error worthy of noting specially, so that particular error
1963 62. When Exim is trying to find all the local interfaces, it used to panic and
1964 die if the ioctl to get the interface flags failed. However, it seems that
1965 on at least one OS (Solaris 9) it is possible to have an interface that is
1966 included in the list of interfaces, but for which you get a failure error
1967 for this call. This happens when the interface is not "plumbed" into a
1968 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1969 failure of the "get flags" call assumes that the interface is down.
1971 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1972 makes life easier for people who are doing arithmetic on fields extracted
1973 from dates, where you often get leading zeros that should not be
1974 interpreted as octal.
1976 64. Added qualify_domain to the redirect router, to override the global
1979 65. If a pathologically long header line contained very many addresses (the
1980 report of this problem mentioned 10 000) and each of them was rewritten,
1981 Exim could use up a very large amount of memory. (It kept on making new
1982 copies of the header line as it rewrote, and never released the old ones.)
1983 At the expense of a bit more processing, the header rewriting function has
1984 been changed so that it no longer eats memory in this way.
1986 66. The generation of the Received: header has been moved from the time that a
1987 message starts to be received, to the time that it finishes. The timestamp
1988 in the Received: header should now be very close to that of the <= log
1989 line. There are two side-effects of this change:
1991 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1992 logged header lines no longer include the local Received: line, because
1993 it has not yet been created. The same applies to a copy of the message
1994 that is returned to a non-SMTP sender when a message is rejected.
1996 (b) When a filter file is tested using -bf, no additional Received: header
1997 is added to the test message. After some thought, I decided that this
2000 This change does not affect the value of $received_for. It is still set
2001 after address rewriting, but before local_scan() is called.
2003 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2005 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2006 gave an unhelpful panic error message, and a defer error. I have managed to
2007 change this behaviour so that it now rejects any supplied certificate,
2008 which seems right, as the list of acceptable certificates is empty.
2010 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2011 gave an unhelpful defer error. I have not managed to make this reject any
2012 supplied certificates, but the error message it gives is "no certificate
2013 supplied", which is not helpful.
2015 70. exigrep's output now also includes lines that are not associated with any
2016 message, but which match the given pattern. Implemented by a patch from
2017 Martin Sluka, which also tidied up the Perl a bit.
2019 71. Recipient callout verification, like sender verification, was using <> in
2020 the MAIL FROM command. This isn't really the right thing, since the actual
2021 sender may affect whether the remote host accepts the recipient or not. I
2022 have changed it to use the actual sender in the callout; this means that
2023 the cache record is now keyed on a recipient/sender pair, not just the
2024 recipient address. There doesn't seem to be a real danger of callout loops,
2025 since a callout by the remote host to check the sender would use <>.
2026 [SEE ABOVE: changed after hitting problems.]
2028 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2029 temporary errors. However, in the case of such a code being given after
2030 the end of a data transmission (i.e. after ".") Exim was failing to write
2031 a retry record for the message. (Yes, there was some broken host that was
2032 actually sending 8xx at this point.)
2034 73. An unknown lookup type in a host list could cause Exim to panic-die when
2035 the list was checked. (An example that provoked this was putting <; in the
2036 middle of a list instead of at the start.) If this happened during a DATA
2037 ACL check, a -D file could be left lying around. This kind of configuration
2038 error no longer causes Exim to die; instead it causes a defer errror. The
2039 incident is still logged to the main and panic logs.
2041 74. Buglet left over from Exim 3 conversion. The message "too many messages
2042 in one connection" was written to the rejectlog but not the mainlog, except
2043 when address rewriting (yes!) was being logged.
2045 75. Added write_rejectlog option.
2047 76. When a system filter was run not as root (that is, when system_filter_user
2048 was set), the values of the $n variables were not being returned to the
2049 main process; thus, they were not subsequently available in the $sn
2052 77. Added +return_path_on_delivery log selector.
2054 78. A connection timeout was being treated differently from recipients deferred
2055 when testing hosts_max_try with a message that was older than the host's
2056 retry timeout. (The host should not be counted, thus allowing all hosts to
2057 be tried at least once before bouncing.) This may have been the cause of an
2058 occasionally reported bug whereby a message would remain on the queue
2059 longer than the retry timeout, but would be bounced if a delivery was
2060 forced. I say "may" because I never totally pinned down the problem;
2061 setting up timeout/retry tests is difficult. See also the next item.
2063 79. The ultimate address timeout was not being applied to errors that involved
2064 a combination of host plus message (for example, a timeout on a MAIL
2065 command). When an address resolved to a number of possible hosts, and they
2066 were not all tried for each delivery (e.g. because of hosts_max_try), a
2067 message could remain on the queue longer than the retry timeout.
2069 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2072 81. Fixed an obscure SMTP outgoing bug which required at least the following
2073 conditions: (a) there was another message waiting for the same server;
2074 (b) the server returned 5xx to all RCPT commands in the first message so
2075 that the message was not completed; (c) the server dropped the connection
2076 or gave a negative response to the RSET that Exim sends to abort the
2077 transaction. The observed case was a dropped connection after DATA that had
2078 been sent in pipelining mode. That is, the server had advertised PIPELINING
2079 but was not implementing it correctly. The effect of the bug was incorrect
2080 behaviour, such as trying another host, and this could lead to a crash.
2086 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2087 and daemon.c were passed as pointers to ints; they should have been
2088 pointers to socklen_t variables (which are typically unsigned ints).
2090 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2093 3. Fixed a really odd bug that affected only the testing scheme; patching a
2094 certain fixed string in the binary changed the value of another string that
2095 happened to be identical to the end of the original first string.
2097 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2098 name", it returns that address as the IP address. On some operating
2099 systems (e.g. Solaris), it also passes back the IP address string as the
2100 "host name". However, on others (e.g. Linux), it passes back an empty
2101 string. Exim wasn't checking for this, and was changing the host name to an
2102 empty string, assuming it had been canonicized.
2104 5. Although rare, it is permitted to have more than one PTR record for a given
2105 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2106 all the names associated with an address, because they do in Solaris.
2107 However, it seems that they do not in Linux for data that comes from the
2108 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2109 I found this out when I moved to a new Linux workstation and tried to run
2110 the Exim test suite.
2112 To get round this problem I have changed the code so that it now does its
2113 own call to the DNS to look up PTR records when searching for a host name.
2114 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2115 addresses that are only in /etc/hosts are still found.
2117 This behaviour is, however, controlled by an option called host_lookup_
2118 order, which defaults to "bydns:byaddr". If people want to use the other
2119 order, or indeed, just use one or the other means of lookup, they can
2120 specify it in this variable.
2122 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2123 some operating systems, this comes back from gethostbyaddr() as an empty
2124 string, and this is what Exim used to test for. However, it seems that in
2125 other systems, "." is yielded. Exim now tests for this case too.
2127 7. The values of check_spool_space and check_log_space are now held internally
2128 as a number of kilobytes instead of an absolute number of bytes. If a
2129 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2130 kilobyte. This means that much larger values can be stored.
2132 8. Exim monitor: an attempt to get the action menu when not actually pointing
2133 at a message produces an empty menu entitled "No message selected". This
2134 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2135 no entries in it ("Shell widget menu has zero width and/or height"). So I
2136 have added a single, blank menu entry in this case.
2138 9. Added ${quote_local_part.
2140 10. MIME decoding is now applied to the contents of Subject: header lines when
2143 11. Now that a reference to $sender_host_address automatically causes a reverse
2144 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2145 host lookup before query-style lookups in lists that might use this
2146 variable. This has therefore been abolished, and the "net-" prefix is no
2147 longer necessary for query-style lookups.
2149 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2150 have been a typo for LFLAGS, so it has been changed.
2152 13. The install script calls Exim with "-C /dev/null" in order to find the
2153 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2154 to be output. Howeve, since Exim outputs its version number before the
2155 error, it didn't break the script. It just looked ugly. I fixed this by
2156 always allowing "-C /dev/null" if the caller is root.
2158 14. Ignore overlarge ACL variable number when reading spool file - insurance
2159 against a later release with more variables having written the file.
2161 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2162 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2163 this, as well as the form without the "IPv6" on the front.
2165 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2166 OS/Makefile-Darwin file.
2168 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2169 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2171 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2172 variable. However, some OS use ints for inodes. I've added cast to long int
2173 to get rid of the compiler warning.
2175 19. I had forgotten to lock out "/../" in configuration file names when
2176 ALT_CONFIG_PREFIX was set.
2178 20. Routers used for verification do not need to specify transports. However,
2179 if such a router generated a host list, and callout was configured, Exim
2180 crashed, because it could not find a port number from the (non-existent)
2181 transport. It now assumes port 25 in this circumstance.
2183 21. Added the -t option to exigrep.
2185 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2186 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2187 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2188 them is set, LOOKUP_LSEARCH is forced.
2190 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2192 24. Added sender and host information to the "rejected by local_scan()" log
2193 line; previously there was no indication of these.
2195 25. Added .include_if_exists.
2197 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2198 Linux. It turns out that not all file systems support this. Apparently some
2199 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2200 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2201 not supported on the file descriptor, is now ignored when Exim is trying to
2202 sync a directory. This applies only to Linux.
2204 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2206 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2207 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2210 29. Added support for maildirsize files from supplied patch (modified a bit).
2212 30. The use of :fail: followed by an empty string could lead Exim to respond to
2213 sender verification failures with (e.g.):
2215 550 Verification failed for <xxx>
2216 550 Sender verify failed
2218 where the first response line was missing the '-' that indicates it is not
2219 the final line of the response.
2221 31. The loop for finding the name of the user that called Exim had a hardwired
2222 limit of 10; it now uses the value of finduser_retries, which is used for
2223 all other user lookups.
2225 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2228 33. Exim was neglecting to zero errno before one call of strtol() when
2229 expanding a string and expecting an integer value. On some systems this
2230 resulted in spurious "integer overflow" errors. Also, it was casting the
2231 result into an int without checking.
2233 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2234 did not work. The code looks as if it has *never* worked, though it appears
2235 to have been documented since at least releast 1.62. I have made it work.
2237 35. The "timeout_DNS" error in retry rules, also documented since at least
2238 1.62, also never worked. As it isn't clear exactly what this means, and
2239 clearly it isn't a major issue, I have abolished the feature by treating it
2240 as "timeout", and writing a warning to the main and panic logs.
2242 36. The display of retry rules for -brt wasn't always showing the error code
2245 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2246 timeout_connect_A, timeout_connect_MX.
2248 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2249 to the empty sender.
2251 39. The daemon was not analysing the content of -oX till after it had closed
2252 stderr and disconnected from the controlling terminal. This meant that any
2253 syntax errors were only noted on the panic log, and the return code from
2254 the command was 0. By re-arranging the code a little, I've made the
2255 decoding happen first, so such errors now appear on stderr, and the return
2256 code is 1. However, the actual setting up of the sockets still happens in
2257 the disconnected process, so errors there are still only recorded on the
2260 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2261 connections (as happens on some IP stacks) was logged at start up time as
2262 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2263 it from "IPv6 and IPv4", which means that two separate sockets are being
2266 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2267 says whether AF_INET or AF_INET6 was passed as an argument.
2269 42. Exiwhat output was messed up when time zones were included in log
2272 43. Exiwhat now gives more information about the daemon's listening ports,
2273 and whether -tls-on-connect was used.
2275 44. The "port" option of the smtp transport is now expanded.
2277 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2278 silently ignored. Now an error message is written to the main and panic
2281 46. There's a new ACL modifier called "logwrite" which writes to a log file
2282 as soon as it is encountered.
2284 47. Added $local_user_uid and $local_user_gid at routing time.
2286 48. Exim crashed when trying to verify a sender address that was being
2289 49. Exim was recognizing only a space character after ".include". It now also
2290 recognizes a tab character.
2292 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2293 extracting the relevant information from the specification. The man page no
2294 longer contains scrambled data for the -d option, and I've added a section
2295 at the front about calling Exim under different names.
2297 51. Added "extra_headers" argument to the "mail" command in filter files.
2299 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2302 53. Installed eximstats 1.29.
2304 54. Added transport_filter_timeout as a generic transport option.
2306 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2307 Cc: header lines. This was required by RFC 822, but it not required by RFC
2310 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2311 incoming messages that did not have them. Now it does so only if the
2312 message originates locally, that is, if there is no associated remote host
2313 address. When Resent- header lines are present, this applies to the Resent-
2314 lines rather than the non-Resent- lines.
2316 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2317 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2319 58. Messages for configuration errors now include the name of the main
2320 configuration file - useful now that there may be more than one file in a
2321 list (.included file names were always shown).
2323 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2324 for those rare installations that do not start the daemon as root or run it
2325 setuid root. I've cut out the call to initgroups() if the daemon is not
2328 60. The Exim user and group can now be bound into the binary as text strings
2329 that are looked up at the start of Exim's processing.
2331 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2333 62. Added $mailstore_basename variable.
2335 63. Installed patch to sieve.c from Michael Haardt.
2337 64. When Exim failed to open the panic log after failing to open the main log,
2338 the original message it was trying to log was written to stderr and debug
2339 output, but if they were not available (the usual case in production), it
2340 was lost. Now it is written to syslog before the two lines that record the
2341 failures to open the logs.
2343 65. Users' Exim filters run in subprocesses under the user's uid. It is
2344 possible for a "deliver" command or an alias in a "personal" command to
2345 provoke an address rewrite. If logging of address rewriting is configured,
2346 this fails because the process is not running as root or exim. There may be
2347 a better way of dealing with this, but for the moment (because 4.30 needs
2348 to be released), I have disabled address rewrite logging when running a
2349 filter in a non-root, non-exim process.
2355 1. The buildconfig auxiliary program wasn't quoting the value set for
2356 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2357 not defined. This bug was masked in 4.22 by the effect that was fixed in
2360 2. Some messages that were rejected after a message id was allocated were
2361 shown as "incomplete" by exigrep. It no longer does this for messages that
2362 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2364 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2365 have allow_domain_literals set, the ID did not get logged in the <= line.
2366 Domain literals are now always recognized in Message-ID: header lines.
2368 4. The first argument for a ${extract expansion item is the key name or field
2369 number. Leading and trailing spaces in this item were not being ignored,
2370 causing some misleading effects.
2372 5. When deliver_drop_privilege was set, single queue runner processes started
2373 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2374 same command in the process it spins off) were not dropping privilege.
2376 6. When the daemon running as "exim" started a queue runner, it always
2377 re-executed Exim in the spun-off process. This is a waste of effort when
2378 deliver_drop_privilege is set. The new process now just calls the
2379 queue-runner function directly.
2385 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2388 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2389 ignored. Though the use of -oP was forcing the writing of a pid file, it
2390 was always written to the default place.
2392 3. If the message "no IP address found for host xxxx" is generated during
2393 incoming verification, it is now followed by identification of the incoming
2394 connection (so you can more easily find what provoked it).
2396 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2398 5. Added some features to "harden" Exim a bit more against certain attacks:
2400 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2401 be put in Local/Makefile. This is like the never_users runtime option,
2402 but it cannot be overridden. The default setting is "root".
2404 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2405 prefix string with which any file named in a -C command line option
2408 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2409 is retained for -C and -D only if the caller of Exim is root. Without
2410 it, the exim user may also use -C and -D and retain privilege.
2412 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2413 command line option is disabled.
2415 6. Macro names set by the -D option must start with an upper case letter, just
2416 like macro names defined in the configuration file.
2418 7. Added "dereference=" facility to LDAP.
2420 8. Two instances of the typo "uknown" in the source files are fixed.
2422 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2423 the Configure-Makefile script screwed up while processing it.
2425 10. Incorporated PCRE 4.4.
2427 11. The SMTP synchronization check was not operating right at the start of an
2428 SMTP session. For example, it could not catch a HELO sent before the client
2429 waited for the greeting. There is now a check for outstanding input at the
2430 point when the greeting is written. Because of the duplex, asynchronous
2431 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2432 way, but not yet received, when the check is performed.
2434 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2435 on TCP/IP sockets, because this apparently causes some broken clients to
2438 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2439 unchanged) from the Cygwin maintainer.
2441 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2442 was supported instead of testing for mbx. Effectively a typo.
2444 15. The spa authenticator server code was not checking that the input it
2445 received was valid base64.
2447 16. The debug output line for the "set" modifier in ACLs was not showing the
2448 name of the variable that was being set.
2450 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2452 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2453 lookup on its own. Something else was needed to trigger the lookup. For
2454 example, a match in host_lookup or the need for a host name in a host list.
2455 Now, if $sender_host_name is referenced and the host name has not yet been
2456 looked up, a lookup is performed. If the lookup fails, the variable remains
2457 empty, and $host_lookup_failed is set to "1".
2459 19. Added "eqi" as a case-independent comparison operator.
2461 20. The saslauthd authentication condition could segfault if neither service
2462 nor realm was specified.
2464 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2465 error message that was logged was misleading, and incoming SMTP
2466 connections were dropped. The message is now more accurate, and temporary
2467 errors are given to SMTP connections.
2469 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2470 (except RSET) with a 421 error, until QUIT is received. However, it was
2471 failing to send a response to QUIT.
2473 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2474 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2475 ACL is now run after the helo_try_verify_hosts code.
2477 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2478 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2479 case-independent, so other case variants are also recognized). Apparently
2480 some systems use these upper case variants.
2482 25. If more than two messages were waiting for the same host, and a transport
2483 filter was specified for the transport, Exim sent two messages over the
2484 same TCP/IP connection, and then failed with "socket operation on non-
2485 socket" when it tried to send the third.
2487 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2489 27. The extern definition of crypt16() in expand.c was not being excluded when
2490 the OS had its own crypt16() function.
2492 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2493 as a preferred synonym for return_size_limit, both as an option and as an
2496 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2498 30. Changed the default configuration ACL to relax the local part checking rule
2499 for addresses that are not in any local domains. For these addresses,
2500 slashes and pipe symbols are allowed within local parts, but the sequence
2501 /../ is explicitly forbidden.
2503 31. SPA server authentication was not clearing the challenge buffer before
2506 32. log_message in a "warn" ACL statement was writing to the reject log as
2507 well as to the main log, which contradicts the documentation and doesn't
2508 seem right (because no rejection is happening). So I have stopped it.
2510 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2511 However, I am unable to do any testing of this.
2513 34. Fixed an infelicity in the appendfile transport. When checking directories
2514 for a mailbox, to see if any needed to be created, it was accidentally
2515 using path names with one or more superfluous leading slashes; tracing
2516 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2518 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2519 contents are added to the log line that is written for every discarded
2520 recipient. (Previously a log_message setting was ignored.)
2522 36. The ${quote: operator now quotes the string if it is empty.
2524 37. The install script runs exim in order to find its version number. If for
2525 some reason other than non-existence or emptiness, which it checks, it
2526 could not run './exim', it was installing it with an empty version number,
2527 i.e. as "exim-". This error state is now caught, and the installation is
2530 38. An argument was missing from the function that creates an error message
2531 when Exim fails to connect to the socket for saslauthd authentication.
2532 This could cause Exim to crash, or give a corrupted message.
2534 39. Added isip, isip4, and isip6 to ${if conditions.
2536 40. The ACL variables $acl_xx are now saved with the message, and can be
2537 accessed later in routers, transports, and filters.
2539 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2540 strings in the file are not string-expanded.
2542 42. If a MAIL command specified a SIZE value that was too large to fit into an
2543 int variable, the check against message_size_limit failed. Such values are
2544 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2545 day this will have to be increased, but I don't think I want to be around
2546 when emails are that large.
2553 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2554 iconv() is not standard in FreeBSD.
2556 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2557 IPv6 enabled. The observed symptom was a segmentation fault on return from
2558 the function os_common_find_running_interfaces() in src/os.c.
2560 3. In the check_special_case() function in daemon.c I had used "errno" as an
2561 argument name, which causes warnings on some systems. This was basically a
2562 typo, since it was named "eno" in the comments!
2564 4. The code that waits for the clock to tick (at a resolution of some fraction
2565 of a second) so as to ensure message-id uniqueness was always waiting for
2566 at least one whole tick, when it could have waited for less. [This is
2567 almost certainly not relevant at current processor speeds, where it is
2568 unlikely to ever wait at all. But we try to future-proof.]
2570 5. The function that sleeps for a time interval that includes fractions of a
2571 second contained a race. It did not block SIGALRM between setting the
2572 timer, and suspending (a couple of lines later). If the interval was short
2573 and the sigsuspend() was delayed until after it had expired, the suspension
2574 never ended. On busy systems this could lead to processes getting stuck for
2577 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2578 process, before it forks any delivery processes. The open lookup caching
2579 mechanism meant that the open file or database connection was passed into
2580 the delivery process. The problem was that delivery processes always tidy
2581 up cached lookup data. This could cause a problem for the next delivery
2582 process started by the queue runner, because the external queue runner
2583 process does not know about the closure. So the next delivery process
2584 still has data in the lookup cache. In the case of a file lookup, there was
2585 no problem because closing a file descriptor in a subprocess doesn't affect
2586 the parent. However, if the lookup was caching a connection to a database,
2587 the connection was closed, and the second delivery process was likely to
2588 see errors such as "PGSQL: query failed: server closed the connection
2589 unexpectedly". The problem has been fixed by closing all cached lookups
2590 in a queue runner before running a delivery process.
2592 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2593 seem to have the "const" qualifier which it has on other OS. I've
2596 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2597 *of the same type* (client or server) with the same public name that an
2598 error should be diagnosed.
2600 9. When Exim looked up a host name for an IP address, but failed to find the
2601 original IP address when looking up the host name (a safety check), it
2602 output the message "<ip address> does not match any IP for NULL", which was
2603 confusing, to say the least. The bug was that the host name should have
2604 appeared instead of "NULL".
2606 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2607 user that is built into the binary, and the -C or -D options is used, root
2608 privilege is dropped before the configuration file is read. In addition,
2609 logging is switched to stderr instead of the normal log files. If the
2610 configuration then re-defines the Exim user, the unprivileged environment
2611 is probably not what is expected, so Exim logs a panic warning message (but
2614 However, if deliver_drop_privilege is set, the unprivileged state may well
2615 be exactly what is intended, so the warning has been cut out in that case,
2616 and Exim is allowed to try to write to its normal log files.
2622 1. smtp_return_error_details was not giving details for temporary sender
2623 or receiver verification errors.
2625 2. Diagnose a configuration error if two authenticators have the same public
2628 3. Exim used not to create the message log file for a message until the first
2629 delivery attempt. This could be confusing when incoming messages were held
2630 for policy or load reasons. The message log file is now created at the time
2631 the message is received, and an initial "Received" line is written to it.
2633 4. The automatically generated man page for command line options had a minor
2634 bug that caused no ill effects; however, a more serious problem was that
2635 the procedure for building the man page automatically didn't always
2636 operate. Consequently, release 4.20 contains an out-of-date version. This
2637 shouldn't happen again.
2639 5. When building Exim with embedded Perl support, the script that builds the
2640 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2641 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2643 6. The freeze_tell option was not being used for messages that were frozen on
2644 arrival, either by an ACL or by local_scan().
2646 7. Added the smtp_incomplete_transaction log selector.
2648 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2649 was accepting AUTH without a new EHLO.
2651 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2652 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2655 10. Logging of TCP/IP connections (when configured) now happens in the main
2656 daemon process instead of the child process, so that the TCP/IP connection
2657 count is more accurate (but it can never be perfect).
2659 11. The use of "drop" in a nested ACL was not being handled correctly in the
2660 outer ACL. Now, if condition failure induced by the nested "drop" causes
2661 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2662 or "require"), the connection is dropped.
2664 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2665 that yield "discard" can now be used with an "accept" or a "discard" verb,
2666 but an error is generated for any others (because I can't see a useful way
2667 to define what should happen).
2669 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2670 are now processed in the same way as lines in the Exim configuration file.
2671 In particular, continuation lines are supported.
2673 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2675 15. Added -ti meaning -t -i.
2677 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2678 domains, and warn by logging if others are found.
2680 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2681 returned from GIFCONF when Exim is trying to find the list of interfaces on
2682 a host. The code in os.c has been modified to copy each ifreq to an aligned
2683 structure in all cases.
2685 Also, in some cases, the returned ifreq's were being copied to a 'struct
2686 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2687 means the last couple of bytes of an IPv6 address could be chopped if the
2688 ifreq contained only a normal sockaddr (14 bytes storage).
2690 18. Named domain lists were not supported in the hosts_treat_as_local option.
2691 An entry such as +xxxx was not recognized, and was treated as a literal
2694 19. Ensure that header lines added by a DATA ACL are included in the reject log
2695 if the ACL subsequently rejects the message.
2697 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2698 MD5 (which is deprecated).
2700 21. When testing a filter file using -bf, Exim was writing a message when it
2701 took the sender from a "From " line in the message, but it was not doing so
2702 when it took $return_path from a Return-Path: header line. It now does.
2704 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2705 with a valid header line field name (a series of printing characters
2706 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2708 23. Changed "disc" in the source to "disk" to conform to the documentation and
2709 the book and for uniformity.
2711 24. Ignore Sendmail's -Ooption=value command line item.
2713 25. When execve() failed while trying to run a command in a pipe transport,
2714 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2715 could be confused with a return value of 69 from the command itself. This
2716 has been changed to 127, the value the shell returns if it is asked to run
2717 a non-existent command. The wording for the related log line suggests a
2718 non-existent command as the problem.
2720 26. If received_header_text expands to an empty string, do not add a Received:
2721 header line to the message. (Well, it adds a token one on the spool, but
2722 marks it "old" so that it doesn't get used or transmitted.)
2724 27. Installed eximstats 1.28 (addition of -nt option).
2726 28. There was no check for failure on the call to getsockname() in the daemon
2727 code. This can fail if there is a shortage of resources on the system, with
2728 ENOMEM, for example. A temporary error is now given on failure.
2730 29. Contrary to the C standard, it seems that in some environments, the
2731 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2732 program. Exim now does this explicitly; it affects the formatting of
2733 timestamps using strftime().
2735 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2736 complaints. I've now initialized all the variables, to avoid this.
2738 32. Header lines added by a system filter were not being "seen" during
2739 transport-time rewrites.
2741 33. The info_callback() function passed to OpenSSL is set up with type void
2742 (*)(SSL *, int, int), as described somewhere. However, when calling the
2743 function (actually a macro) that sets it up, the type void(*)() is
2744 expected. I've put in a cast to prevent warnings from picky compilers.
2746 34. If a DNS black list lookup found a CNAME record, but there were no A
2747 records associated with the domain it pointed at, Exim crashed.
2749 35. If a DNS black list lookup returned more than one A record, Exim ignored
2750 all but the first. It now scans all returned addresses if a particular IP
2751 value is being sought. In this situation, the contents of the
2752 $dnslist_value variable are a list of all the addresses, separated by a
2755 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2756 to accept a host name and all its aliases if the forward lookup for any of
2757 them yielded the IP address of the incoming connection. Now it accepts only
2758 those names whose forward lookup yields the correct IP address. Any other
2759 names are discarded. This closes a loophole whereby a rogue DNS
2760 administrator could create reverse DNS records to break through a
2761 wildcarded host restriction in an ACL.
2763 37. If a user filter or a system filter that ran in a subprocess used any of
2764 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2765 the wrong values were passed to the pipe command ($thisaddress had the
2766 value of $0, $0 had the value of $1, etc). This bug was introduced by
2767 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2769 38. Improved the line breaking for long SMTP error messages from ACLs.
2770 Previously, if there was no break point between 40 and 75 characters, Exim
2771 left the rest of the message alone. Two changes have been made: (a) I've
2772 reduced the minimum length to 35 characters; (b) if it can't find a break
2773 point between 35 and 75 characters, it looks ahead and uses the first one
2774 that it finds. This may give the occasional overlong line, but at least the
2775 remaining text gets split now.
2777 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2778 file descriptors might be low, and that setting 1000 would always raise it.
2779 It turns out that in some environments, the limit is already over 1000 and
2780 that lowering it causes trouble. So now Exim takes care not to decrease it.
2782 40. When delivering a message, the value of $return_path is set to $sender_
2783 address at the start of routing (routers may change the value). By an
2784 oversight, this default was not being set up when an address was tested by
2785 -bt or -bv, which affected the outcome if any router or filter referred to
2788 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2789 log only when "message" or "log_message" are set. However, if one of the
2790 conditions was an address verification, or a call to a nested ACL, the
2791 messages generated by the underlying test were being passed through. This
2792 no longer happens. The underlying message is available in $acl_verify_
2793 message for both "message" and "log_message" expansions, so it can be
2794 passed through if needed.
2796 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2797 new expansion $bh_ to give the encoded byte string without charset
2798 translation. Translation happens only if iconv() is available; HAVE_ICONV
2799 indicates this at build time. HEADERS_CHARSET gives the charset to
2800 translate to; headers_charset can change it in the configuration, and
2801 "headers charset" can change it in an individual filter file.
2803 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2804 that creates RFC 2047 encoded "words" labels them as that charset instead
2805 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2806 expansion operator; (ii) when Exim creates a From: line for a local
2807 message; (iii) when a header line is rewritten to include a "phrase" part.
2809 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2810 buggy, causing it to skip the first lines of messages whose message ID
2811 ended in 'D'. This would not have bitten before Exim release 4.14, because
2812 message IDs were unlikely to end in 'D' before then. The effect was to have
2813 incorrect size information for certain domains.
2815 45. #include "config.h" was missing at the start of the crypt16.c module. This
2816 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2819 46. If there was a timeout during a "random" callout check, Exim treated it as
2820 a failure of the random address, and carried on sending RSET and the real
2821 address. If the delay was just some slowness somewhere, the response to the
2822 original RCPT would be taken as a response to RSET and so on, causing
2823 mayhem of various kinds.
2825 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2826 when I implemented it. It didn't allow for the fact that some option values
2827 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2828 the right test for positive values.
2830 48. Domain names in DNS records are case-independent. Exim always looks them up
2831 in lower case. Some resolvers return domain names in exactly the case they
2832 appear in the zone file, that is, they may contain uppercase letters. Not
2833 all resolvers do this - some return always lower case. Exim was treating a
2834 change of case by a resolver as a change of domain, similar to a widening
2835 of a domain abbreviation. This triggered its re-routing code and so it was
2836 trying to route what was effectively the same domain again. This normally
2837 caused routing to fail (because the router wouldn't handle the domain
2838 twice). Now Exim checks for this case specially, and just changes the
2839 casing of the domain that it ultimately uses when it transmits the message
2842 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2845 50. If a filter generated a file delivery with a non-absolute name (possible if
2846 no home directory exists for the router), the forbid_file option was not
2849 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2850 the existing equality matching.
2852 52. Exim was using ints instead of ino_t variables in some places where it was
2853 dealing with inode numbers.
2855 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2856 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2857 called TMPDIR, and if it finds it is different, it changes its value.
2859 54. The smtp_printf() function is now made available to local_scan() so
2860 additional output lines can be written before returning. There is also an
2861 smtp_fflush() function to enable the detection of a dropped connection.
2862 The variables smtp_input and smtp_batched_input are exported to
2865 55. Changed the default runtime configuration: the message "Unknown user"
2866 has been removed from the ACL, and instead placed on the localuser router,
2867 using the cannot_route_message feature. This means that any verification
2868 failures that generate their own messages won't get overridden. Similarly,
2869 the "Unrouteable address" message that was in the ACL for unverifiable
2870 relay addresses has also been removed.
2872 56. Added hosts_avoid_esmtp to the smtp transport.
2874 57. The exicyclog script was not checking for the esoteric option
2875 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2876 will work only if exicyclog is run under the appropriate euid.
2878 58. Following a discussion on the list, the rules by which Exim recognises line
2879 endings on incoming messages have been changed. The -dropcr and drop_cr
2880 options are now no-ops, retained only for backwards compatibility. The
2881 following line terminators are recognized: LF CRLF CR. However, special
2882 processing applies to CR:
2884 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2885 nor a local message in the state where . is a terminator.
2887 (ii) If a bare CR is encountered in a header line, an extra space is added
2888 after the line terminator so as not to end the header. The reasoning
2889 behind this is that bare CRs in header lines are most likely either
2890 to be mistakes, or people trying to play silly games.
2892 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2893 was being incorrectly given as 18 bytes larger than it should have been.
2894 This is a VOB (very old bug).
2896 60. This may never have affected anything current, but just in case it has:
2897 When the local host is found other than at the start of a list of hosts,
2898 the local host, those with the same MX, and any that follow, are discarded.
2899 When the list in question was part of a longer list of hosts, the following
2900 hosts (not currently being processed) were also being discarded. This no
2901 longer happens. I'm not sure if this situation could ever has previously
2904 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2907 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2908 header. This is something that is recommended in a new Internet Draft, and
2909 is something that is documented as being done by Sendmail. There are two
2910 possible values. For messages generated by the autoreply transport, Exim
2913 Auto-Submitted: auto-replied
2915 whereas for all other generated messages (e.g. bounces) it adds
2917 Auto-Submitted: auto-generated
2919 63. The "personal" condition in filters now includes a test for the
2920 Auto-Submitted: header. If it contains the string "auto-" the message it
2921 not considered personal.
2923 64. Added rcpt_include_affixes as a generic transport option.
2925 65. Added queue_only_override (default true).
2927 66. Added the syslog_duplication option.
2929 67. If what should have been the first header line of a message consisted of
2930 a space followed by a colon, Exim was mis-interpreting it as a header line.
2931 It isn't of course - it is syntactically invalid and should therefore be
2932 treated as the start of the message body. The misbehaviour could have
2933 caused a number of strange effects, including loss of data in subsequent
2934 header lines, and spool format errors.
2936 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2937 client host had authenticated. This control can now be exercised by an ACL
2938 for more flexibility.
2940 69. By default, callouts do not happen when testing with -bh. There is now a
2941 variant, -bhc, which does actually run the callout code, including
2942 consulting and updating the callout cache.
2944 70. Added support for saslauthd authentication, courtesy of Alexander
2947 71. If statvfs() failed on the spool or log directories while checking their
2948 size for availability, Exim confusingly gave the error "space shortage".
2949 Furthermore, in debugging mode it crashed with a floating point exception.
2950 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2951 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2952 serious problem, Exim now writes to the main and panic logs when this
2953 happens, with details of the failure. It then refuses to accept the
2954 incoming message, giving the message "spool directory problem" or "log
2955 directory problem" with a 421 code for SMTP messages.
2957 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2958 0, 1, and 2 exist, because some OS complain for execs without them (see
2959 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2960 descriptors. However, the code omitted to check that the open succeeded,
2961 causing mysterious errors if for some reason the permissions on /dev/null
2962 got screwed. Now Exim writes a message to the main and panic logs, and
2963 bombs out if it can't open /dev/null.
2965 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2966 interact so that it is all more flexible. It is supposed to remain
2967 backwards compatible. Also added extra_local_interfaces.
2969 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2970 to bomb out with an assertion failure - to the client this appears as a
2971 connection drop. This problem occurs in the part of the code that was taken
2972 from the Samba project. Fortunately, the assertion is in a very simple
2973 function, so I have fixed this by reproducing the function inline in the
2974 one place where it is called, and arranging for authentication to fail
2975 instead of killing the process with assert().
2977 75. The SPA client code was not working when the server requested OEM rather
2978 than Unicode encoding.
2980 76. Added code to make require_files with a specific uid setting more usable in
2981 the case where statting the file as root fails - usually a non-root-mounted
2982 NFS file system. When this happens and the failure is EACCES, Exim now
2983 forks a subprocess and does the per-uid checking as the relevant uid.
2985 77. Added process_log_path.
2987 78. If log_file_path was not explicitly set, a setting of check_log_space or
2988 check_log_inodes was ignored.
2990 79. If a space check for the spool or log partitions fails, the incident is now
2991 logged. Of course, in the latter case the data may get lost...
2993 80. Added the %p formatting code to string_format() so that it can be used to
2994 print addresses in debug_print(). Adjusted all the address printing in the
2995 debugging in store.c to use %p rather than %d.
2997 81. There was a concern that a line of code in smtp_in.c could overflow a
2998 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2999 initially expressed, the concern was not well-founded, because trailing
3000 spaces are removed early. However, if the trailing spaces were followed by
3001 a NULL, they did not get removed, so the overflow was possible. Two fixes
3004 (a) I re-wrote the offending code in a cleaner fashion.
3005 (b) If an incoming SMTP command contains a NULL character, it is rejected
3008 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3009 runs initgroups(), so that if the Exim user is in any additional groups,
3010 they will be used during message reception.
3016 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob