1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.152 2005/06/10 19:27:05 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
12 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
13 command itself ended in error, the underlying message about the transport
14 filter timeout was being overwritten with the pipe command error. Now the
15 underlying error message should be appended to the second error message.
17 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
19 PH/02 Reduce the amount of output that "make" produces by default. Full output
20 can still be requested.
22 PH/03 The warning log line about a condition test deferring for a "warn" verb
23 was being output only once per connection, rather than after each
24 occurrence (because it was using the same function as for successful
25 "warn" verbs). This seems wrong, so I have changed it.
27 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
28 it should not have, which might have caused a crash in the right
29 circumstances, but probably never did.
31 PH/04 Installed a modified version of Tony Finch's patch to make submission
32 mode fix the return path as well as the Sender: header line, and to
33 add a /name= option so that you can make the user's friendly name appear
36 TF/03 Added the control = fakedefer ACL modifier.
38 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
39 Mark Lowes for thorough testing.
41 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
43 TK/03 Merged latest SRS patch from Miles Wilton.
45 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
46 with the definition in sysexits.h (which is #included earlier).
47 Fortunately, Exim does not actually use EX_OK. The code used to try to
48 preserve the sysexits.h value, by assumimg that macro definitions were
49 scanned for macro replacements. I have been disabused of this notion,
50 so now the code just undefines EX_OK before #including unistd.h.
52 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
53 in the smtp transport. When a block could not be written in a single
54 write() function, the timeout was being re-applied to each part-write.
55 This seems wrong - if the receiver was accepting one byte at a time it
56 would take for ever. The timeout is now adjusted when this happens. It
57 doesn't have to be particularly precise.
59 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
60 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
62 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
63 without a domain if the check to <postmaster@domain> fails.
65 SC/01 Eximstats: added -xls and the ability to specify output files
66 (patch written by Frank Heydlauf).
68 SC/02 Eximstats: use FileHandles for outputing results.
70 SC/03 Eximstats: allow any combination of xls, txt, and html output.
72 SC/04 Eximstats: fixed display of large numbers with -nvr option
74 SC/05 Eximstats: fixed merging of reports with empty tables.
76 SC/06 Eximstats: added the -include_original_destination flag
78 SC/07 Eximstats: removed tabs and trailing whitespace.
80 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
82 TK/06 MBOX spool code: Add real "From " MBOX separator line
83 so the .eml file is really in mbox format (even though
84 most programs do not really care). Patch from Alex Miller.
86 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
87 The latter is generated from $received_to and is only set if the
88 message has one envelope recipient. SA can use these headers,
89 obviously out-of-the-box. Patch from Alex Miller.
91 PH/08 The ${def test on a variable was returning false if the variable's
92 value was "0", contrary to what the specification has always said!
93 The result should be true unless the variable is empty.
95 PH/09 The syntax error of a character other than { following "${if
96 def:variable_name" (after optional whitespace) was not being diagnosed.
97 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
98 accidental colon was present, for example, could give incorrect results.
100 PH/10 Tidied the code in a number of places where the st_size field of a stat()
101 result is used (not including appendfile, where other changes are about
104 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
105 This involved changing a lot of size variables from int to off_t. It
106 should work with maildirs and everything.
108 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
109 spamd dying while we are connected to it.
115 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
116 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
118 TK/02 Fix ACL "control" statement not being available in MIME ACL.
120 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
122 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
123 to test Sieve filters that use "vacation".
125 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
126 that changes the way the GnuTLS parameters are stored in the cache file.
127 The new format can be generated externally. For backward compatibility,
128 if the data in the cache doesn't make sense, Exim assumes it has read an
129 old-format file, and it generates new data and writes a new file. This
130 means that you can't go back to an older release without removing the
133 PH/03 A redirect router that has both "unseen" and "one_time" set does not
134 work if there are any delivery delays because "one_time" forces the
135 parent to be marked "delivered", so its unseen clone is never tried
136 again. For this reason, Exim now forbids the simultaneous setting of
139 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
140 redirected to themselves ("homonym" addresses). Read the long ChangeLog
141 entry if you want to know the details. The fix, however, neglected to
142 consider the case when local delivery batching is involved. The test for
143 "previously delivered" was not happening when checking to see if an
144 address could be batched with a previous (undelivered) one; under
145 certain circumstances this could lead to multiple deliveries to the same
148 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
149 in its include files, and this causes problems building Exim.
151 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
152 header_syntax) but Exim was just ignoring anything given after a slash.
153 In particular, this caused confusion with an attempt to use "verify =
154 reverse_host_lookup/defer_ok". An error is now given when options are
155 supplied for verify items that do not have them. (Maybe reverse_host_
156 lookup should have a defer_ok option, but that's a different point.)
158 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
159 defined by RFC 821) to 2048, because there were problems with some AUTH
160 commands, and RFC 1869 says the size should be increased for extended
161 SMTP commands that take arguments.
163 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
166 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
167 "unknown" error; now it says that the functionality isn't in the binary.
169 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
170 an address' error message when a string expansion fails (syntax or
171 whatever). Otherwise the password may appear in the log. Following change
172 PH/42 below, there is no longer a chance of it appearing in a bounce
175 PH/11 Installed exipick version 20050225.0 from John Jetmore.
177 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
178 of its addresses was ever tried. (Bugzilla bug #2.)
180 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
181 the result incorrectly in the debug output. (It correctly added a newline
182 to what was transported.)
184 TF/01 Added $received_time.
186 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
187 commented out examples of how to interface to a virus scanner and to
188 SpamAssassin. Also added commented examples of av_scanner and
189 spamd_address settings.
191 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
192 and controls are allowed in which ACLs. There were a couple of minor
193 errors. Some of the entries in the conditions table (which is a table of
194 where they are NOT allowed) were getting very unwieldy; rewrote them as a
195 negation of where the condition IS allowed.
197 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
199 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
200 header file does not have a version number, so I've had to invent a new
201 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
202 API. The code is untested by me (my Linux distribution still has 0.3.2 of
203 radiusclient), but it was contributed by a Radius user.
205 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
206 files or directories, for OpenSSL.
208 PH/19 When an Exim process that is running as root has to create an Exim log
209 file, it does so in a subprocess that runs as exim:exim so as to get the
210 ownership right at creation (otherwise, other Exim processes might see
211 the file with the wrong ownership). There was no test for failure of this
212 fork() call, which would lead to the process getting stuck as it waited
213 for a non-existent subprocess. Forks do occasionally fail when resources
214 run out. I reviewed all the other calls to fork(); they all seem to check
217 PH/20 When checking for unexpected SMTP input at connect time (before writing
218 the banner), Exim was not dealing correctly with a non-positive return
219 from the read() function. If the client had disconnected by this time,
220 the result was a log entry for a synchronization error with an empty
221 string after "input=" when read() returned zero. If read() returned -1
222 (an event I could not check), uninitialized data bytes were printed.
223 There were reports of junk text (parts of files, etc) appearing after
226 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
228 PH/22 Added support for macro redefinition, and (re)definition in between
229 driver and ACL definitions.
231 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
232 forgetting to use the resulting value; it was using the unexpanded value.
234 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
235 hadn't been configured. The fix is from Juergen Kreileder, who
236 understands it better than I do:
238 "Here's what I see happening with three configured cyrus_sasl
239 authenticators configured (plain, login, cram-md5):
241 On startup auth_cyrus_sasl_init() gets called for each of these.
242 This means three calls to sasl_listmech() without a specified mech_list.
243 => SASL tests which mechs of all available mechs actually work
244 => three warnings about OTP not working
245 => the returned list contains: plain, login, cram-md5, digest-md5, ...
247 With the patch, sasl_listmech() also gets called three times. But now
248 SASL's mech_list option is set to the server_mech specified in the the
249 authenticator. Or in other words, the answer from sasl_listmech()
250 gets limited to just the mech you're testing for (which is different
252 => the return list contains just 'plain' or 'login', 'cram-md5' or
253 nothing depending on the value of ob->server_mech.
255 I've just tested the patch: Authentication still works fine,
256 unavailable mechs specified in the exim configuration are still
257 caught, and the auth.log warnings about OTP are gone."
259 PH/25 When debugging is enabled, the contents of the command line are added
260 to the debugging output, even when log_selector=+arguments is not
263 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
264 answer is "GNU", and only if the return is "GNU/something" is the answer
267 PH/27 $acl_verify_message is now set immediately after the failure of a
268 verification in an ACL, and so is available in subsequent modifiers. In
269 particular, the message can be preserved by coding like this:
271 warn !verify = sender
272 set acl_m0 = $acl_verify_message
274 Previously, $acl_verify_message was set only while expanding "message"
275 and "log_message" when a very denied access.
277 PH/28 Modified OS/os.c-Linux with
279 -#ifndef OS_LOAD_AVERAGE
280 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
282 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
283 nomenclature these days.)
285 PH/29 Installed patch from the Sieve maintainer that adds the options
286 sieve_useraddress and sieve_subaddress to the redirect router.
288 PH/30 In these circumstances:
289 . Two addresses routed to the same list of hosts;
290 . First host does not offer TLS;
291 . First host accepts first address;
292 . First host gives temporary error to second address;
293 . Second host offers TLS and a TLS session is established;
294 . Second host accepts second address.
295 Exim incorrectly logged both deliveries with the TLS parameters (cipher
296 and peerdn, if requested) that were in fact used only for the second
299 PH/31 When doing a callout as part of verifying an address, Exim was not paying
300 attention to any local part prefix or suffix that was matched by the
301 router that accepted the address. It now behaves in the same way as it
302 does for delivery: the affixes are removed from the local part unless
303 rcpt_include_affixes is set on the transport.
305 PH/32 Add the sender address, as F=<...>, to the log line when logging a
306 timeout during the DATA phase of an incoming message.
308 PH/33 Sieve envelope tests were broken for match types other than :is. I have
309 applied a patch sanctioned by the Sieve maintainer.
311 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
312 the uid or gid is negative. A case of a negative gid caused this to be
313 noticed. The fix allows for either to be negative.
315 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
316 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
319 PH/36 The change PH/12 above was broken. Fixed it.
321 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
322 the grounds that routing the same address twice would always produce the
323 same answer. This might have been true once, but it is certainly no
324 longer true now. Routing a child address may depend on the previous
325 routing that produced that child. Some complicated redirection strategies
326 went wrong when messages had multiple recipients, and made Exim's
327 behaviour dependent on the order in which the addresses were given.
329 I have moved the duplicate checking until after the routing is complete.
330 Exim scans the addresses that are assigned to local and remote
331 transports, and removes any duplicates. This means that more work will be
332 done, as duplicates will always all be routed, but duplicates are
333 presumably rare, so I don't expect this is of any significance.
335 For deliveries to pipes, files, and autoreplies, the duplicate checking
336 still happens during the routing process, since they are not going to be
339 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
340 It corrects a timeout issue with spamd. This is Ian's comment: "The
341 background is that sometimes spamd either never reads data from a
342 connection it has accepted, or it never writes response data. The exiscan
343 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
344 blindly assumes that writes won't block so it may never time out."
346 PH/39 Allow G after quota size as well as K and M.
348 PH/40 The value set for $authenticated_id in an authenticator may not contain
349 binary zeroes or newlines because the value is written to log lines and
350 to spool files. There was no check on this. Now the value is run through
351 the string_printing() function so that such characters are converted to
352 printable escape sequences.
354 PH/41 $message_linecount is a new variable that contains the total number of
355 lines in the message. Compare $body_linecount, which is the count for the
358 PH/42 Exim no longer gives details of delivery errors for specific addresses in
359 bounce and delay warning messages, except in certain special cases, which
362 (a) An SMTP error message from a remote host;
363 (b) A message specified in a :fail: redirection;
364 (c) A message specified in a "fail" command in a system filter;
365 (d) A message specified in a FAIL return from the queryprogram router;
366 (e) A message specified by the cannot_route_message router option.
368 In these cases only, Exim does include the error details in bounce and
369 warning messages. There are also a few cases where bland messages such
370 as "unrouteable address" or "local delivery error" are given.
372 PH/43 $value is now also set for the "else" part of a ${run expansion.
374 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
375 being worked on, but at least Exim now implements the latest version to
378 PH/45 In a pipe transport, although a timeout while waiting for the pipe
379 process to complete was treated as a delivery failure, a timeout while
380 writing the message to the pipe was logged, but erroneously treated as a
381 successful delivery. Such timeouts include transport filter timeouts. For
382 consistency with the overall process timeout, these timeouts are now
383 treated as errors, giving rise to delivery failures by default. However,
384 there is now a new Boolean option for the pipe transport called
385 timeout_defer, which, if set TRUE, converts the failures into defers for
386 both kinds of timeout. A transport filter timeout is now identified in
389 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
390 systems where "make" and "gmake" are different, calling "gmake" at top
391 level broke things. I've arranged for the value of $(MAKE) to be passed
392 from the Makefile to this script so that it can call the same version of
396 A note about Exim versions 4.44 and 4.50
397 ----------------------------------------
399 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
400 changes of various kinds. As a consequence, a big documentation update was
401 needed. This delayed the release for rather longer than seemed good, especially
402 in the light of a couple of (minor) security issues. Therefore, the changes
403 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
404 release. So 4.44 and 4.50 are in effect two different branches that both start
407 I have left the 4.50 change log unchanged; it contains all the changes since
408 4.43. The change log for 4.44 is below; many of its items are identical to
409 those for 4.50. This seems to be the most sensible way to preserve the
410 historical information.
416 1. Minor wording change to the doc/README.SIEVE file.
418 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
419 computation of the current number of files was incorrect.
421 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
422 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
425 4. Give more explanation in the error message when the command for a transport
426 filter fails to execute.
428 5. There are several places where Exim runs a non-Exim command in a
429 subprocess. The SIGUSR1 signal should be disabled for these processes. This
430 was being done only for the command run by the queryprogram router. It is
431 now done for all such subprocesses. The other cases are: ${run, transport
432 filters, and the commands run by the lmtp and pipe transports.
434 6. Added CONFIGURE_GROUP build-time option.
436 7. Some older OS have a limit of 256 on the maximum number of file
437 descriptors. Exim was using setrlimit() to set 1000 as a large value
438 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
439 systems. I've change it so that if it can't get 1000, it tries for 256.
441 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
442 was an oversight, and furthermore, ever since the addition of extra
443 controls (e.g. 4.43/32), the checks on when to allow different forms of
444 "control" were broken. There should now be diagnostics for all cases when a
445 control that does not make sense is encountered.
447 9. Added the /retain_sender option to "control=submission".
449 10. $recipients is now available in the predata ACL (oversight).
451 11. Tidy the search cache before the fork to do a delivery from a message
452 received from the command line. Otherwise the child will trigger a lookup
453 failure and thereby defer the delivery if it tries to use (for example) a
454 cached ldap connection that the parent has called unbind on.
456 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
457 of $address_data from the recipient verification was clobbered by the
460 13. The value of address_data from a sender verification is now available in
461 $sender_address_data in subsequent conditions in the ACL statement.
463 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
465 15. Added a new option "connect=<time>" to callout options, to set a different
468 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
469 was its contents. (It was OK if the option was not defined at all.)
471 17. A "Completed" log line is now written for messages that are removed from
472 the spool by the -Mrm option.
474 18. New variables $sender_verify_failure and $recipient_verify_failure contain
475 information about exactly what failed.
477 19. Added -dd to debug only the daemon process.
479 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
480 handles timeouts, both on the server side and network timeouts. Renamed the
481 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
483 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
484 It is now set to "smtps".
486 22. $host_address is now set to the target address during the checking of
489 23. When checking ignore_target_hosts for an ipliteral router, no host name was
490 being passed; this would have caused $sender_host_name to have been used if
491 matching the list had actually called for a host name (not very likely,
492 since this list is usually IP addresses). A host name is now passed as
495 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
496 code that specifies a non-restarting handler (typically sigaction() in
497 modern systems) in an attempt to fix a rare and obscure crash bug.
499 25. Narrowed the window for a race in the daemon that could cause it to ignore
500 SIGCHLD signals. This is not a major problem, because they are used only to
501 wake it up if nothing else does.
503 26. A malformed maildirsize file could cause Exim to calculate negative values
504 for the mailbox size or file count. Odd effects could occur as a result.
505 The maildirsize information is now recalculated if the size or filecount
508 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
509 support for a long time. Removed HAVE_SYS_VFS_H.
511 28. Installed the latest version of exipick from John Jetmore.
513 29. In an address list, if the pattern was not a regular expression, an empty
514 subject address (from a bounce message) matched only if the pattern was an
515 empty string. Non-empty patterns were not even tested. This was the wrong
516 because it is perfectly reasonable to use an empty address as part of a
517 database query. An empty address is now tested by patterns that are
518 lookups. However, all the other forms of pattern expect the subject to
519 contain a local part and a domain, and therefore, for them, an empty
520 address still always fails if the pattern is not itself empty.
522 30. Exim went into a mad DNS loop when attempting to do a callout where the
523 host was specified on an smtp transport, and looking it up yielded more
526 31. Re-factored the code for checking spool and log partition space into a
527 function that finds that data and another that does the check. The former
528 is then used to implement four new variables: $spool_space, $log_space,
529 $spool_inodes, and $log_inodes.
531 32. The RFC2047 encoding function was originally intended for short strings
532 such as real names; it was not keeping to the 75-character limit for
533 encoded words that the RFC imposes. It now respects the limit, and
534 generates multiple encoded words if necessary. To be on the safe side, I
535 have increased the buffer size for the ${rfc2047: expansion operator from
538 33. It is now permitted to omit both strings after an "if" condition; if the
539 condition is true, the result is "true". As before, when the second string
540 is omitted, a false condition yields an empty string. This makes it less
541 cumbersome to write custom ACL and router conditions.
543 34. Failure to deliver a bounce message always caused it to be frozen, even if
544 there was an errors_to setting on the router. The errors_to setting is now
547 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
548 canonical form (fully expanded) before being placed in
549 $sender_host_address.
551 36. The table in the code that translates DNS record types into text (T_A to
552 "A" for instance) was missing entries for NS and CNAME. It is just possible
553 that this could have caused confusion if both these types were looked up
554 for the same domain, because the text type is used as part of Exim's
555 per-process caching. But the chance of anyone hitting this buglet seems
558 37. The dnsdb lookup has been extended in a number of ways.
560 (1) There is a new type, "zns", which walks up the domain tree until it
561 finds some nameserver records. It should be used with care.
563 (2) There is a new type, "mxh", which is like "mx" except that it returns
564 just the host names, not the priorities.
566 (3) It is now possible to give a list of domains (or IP addresses) to be
567 looked up. The behaviour when one of the lookups defers can be
568 controlled by a keyword.
570 (4) It is now possible to specify the separator character for use when
571 multiple records are returned.
573 38. The dnslists ACL condition has been extended: it is now possible to supply
574 a list of IP addresses and/or domains to be looked up in a particular DNS
577 39. Added log_selector=+queue_time_overall.
579 40. When running the queue in the test harness, wait just a tad after forking a
580 delivery process, to get repeatability of debugging output.
582 41. Include certificate and key file names in error message when GnuTLS fails
583 to set them up, because the GnuTLS error message doesn't include the name
584 of the failing file when there is a problem reading it.
586 42. Allow both -bf and -bF in the same test run.
588 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
590 44. The "Exiscan patch" is now merged into the mainline Exim source.
592 45. Sometimes the final signoff response after QUIT could fail to get
593 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
594 < 0 before doing a fflush(). This bug looks as though it goes back to the
595 introduction of TLS in release 3.20, but "sometimes" must have been rare
596 because the tests only now provoked it.
598 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
599 (this can affect the format of dates).
601 47. exim_tidydb, when checking for the continued existence of a message for
602 which it has found a message-specific retry record, was not finding
603 messages that were in split spool directories. Consequently, it was
604 deleting retry records that should have stayed in existence.
606 48. Steve fixed some bugs in eximstats.
608 49. The SPA authentication driver was not abandoning authentication and moving
609 on to the next authenticator when an expansion was forced to fail,
610 contradicting the general specification for all authenticators. Instead it
611 was generating a temporary error. It now behaves as specified.
613 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
614 (the order specifies the preference for clients). The order is now AES256,
615 AES128, 3DES, ARCFOUR128.
617 51. Small patch to Sieve code - explicitly set From: when generating an
620 52. Exim crashed if a remote delivery caused a very long error message to be
621 recorded - for instance if somebody sent an entire SpamAssassin report back
622 as a large number of 550 error lines. This bug was coincidentally fixed by
623 increasing the size of one of Exim's internal buffers (big_buffer) that
624 happened as part of the Exiscan merge. However, to be on the safe side, I
625 have made the code more robust (and fixed the comments that describe what
628 53. Now that there can be additional text after "Completed" in log lines (if
629 the queue_time_overall log selector is set), a one-byte patch to exigrep
630 was needed to allow it to recognize "Completed" as not the last thing in
633 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
634 patch that reportedly fixes this has been added. I am not expert enough to
635 create a test for it. This is what the patch creator wrote:
637 "I found a little strange behaviour of ldap code when working with
638 Windows 2003 AD Domain, where users was placed in more than one
639 Organization Units. When I tried to give exim partial DN, the exit code
640 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
641 But simultaneously result of request was absolutely normal ldap result,
642 so I produce this patch..."
644 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
645 REFERENCE, so I have modified the code to exclude the patch when that macro
648 55. Some experimental protocols are using DNS PTR records for new purposes. The
649 keys for these records are domain names, not reversed IP addresses. The
650 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
651 leaves it alone. Component reversal etc. now happens only for IP addresses.
654 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
656 57. Double the size of the debug message buffer (to 2048) so that more of very
657 long debug lines gets shown.
659 58. The exicyclog utility now does better if the number of log files to keep
660 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
662 59. Two changes related to the smtp_active_hostname option:
664 (1) $smtp_active_hostname is now available as a variable.
665 (2) The default for smtp_banner uses $smtp_active_hostname instead
666 of $primary_hostname.
668 60. The host_aton() function is supposed to be passed a string that is known
669 to be a valid IP address. However, in the case of IPv6 addresses, it was
670 not checking this. This is a hostage to fortune. Exim now panics and dies
671 if the condition is not met. A case was found where this could be provoked
672 from a dnsdb PTR lookup with an IPv6 address that had more than 8
673 components; fortuitously, this particular loophole had already been fixed
674 by change 4.50/55 above.
676 If there are any other similar loopholes, the new check in host_aton()
677 itself should stop them being exploited. The report I received stated that
678 data on the command line could provoke the exploit when Exim was running as
679 exim, but did not say which command line option was involved. All I could
680 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
684 61. There was a buffer overflow vulnerability in the SPA authentication code
685 (which came originally from the Samba project). I have added a test to the
686 spa_base64_to_bits() function which I hope fixes it.
689 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
690 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
692 63. The daemon start-up calls getloadavg() while still root for those OS that
693 need the first call to be done as root, but it missed one case: when
694 deliver_queue_load_max is set with deliver_drop_privilege. This is
695 necessary for the benefit of the queue runner, because there is no re-exec
696 when deliver_drop_privilege is set.
698 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
701 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
702 was in use, was not putting the data itself into the right store pool;
703 consequently, it could be overwritten for a subsequent message in the same
704 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
707 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
709 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
710 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
711 the function were treating the return as a boolean value, which happened to
712 work because 0=false and not-0=true, but is not correct code.
714 68. The host_aton() function was not handling scoped IPv6 addresses (those
715 with, for example, "%eth0" on the end) correctly.
717 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
718 negated items (that is, ~something) in unsigned ints. Some compilers
719 apparently mutter when there is no cast.
721 70. If an address verification called from an ACL failed, and did not produce a
722 user-specific message (i.e. there was only a "system" message), nothing was
723 put in $acl_verify_message. In this situation, it now puts the system
726 71. Change 4.23/11 added synchronization checking at the start of an SMTP
727 session; change 4.31/43 added the unwanted input to the log line - except
728 that it did not do this in the start of session case. It now does.
730 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
731 This is wrong and can cause the other end to generate a synchronization
732 error if it is another Exim or anything else that does the synchronization
733 check. A QUIT command is no longer sent after a timeout.
735 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
738 74. The defer_ok option of callout verification was not working if it was used
739 when verifying addresses in header lines, that is, for this case:
741 verify = header_sender/callout=defer_ok
743 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
744 those file descriptors could be used for SMTP connections. If anything
745 wrote to stderr (the example that came up was "warn" in embedded Perl), it
746 could be sent to the SMTP client, causing chaos. The daemon now opens
747 stdin, stdout, and stderr to /dev/null when it puts itself into the
750 76. Arrange for output from Perl's "warn" command to be written to Exim's main
751 log by default. The user can override this with suitable Perl magic.
753 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
754 the log message when discard triggers, was not working for the DATA ACL or
755 for the non-SMTP ACL.
757 78. Error message wording change in sieve.c.
759 79. If smtp_accept_max_per_host was set, the number of connections could be
760 restricted to fewer than expected, because the daemon was trying to set up
761 a new connection before checking whether the processes handling previous
762 connections had finished. The check for completed processes is now done
763 earlier. On busy systems, this bug wouldn't be noticed because something
764 else would have woken the daemon, and it would have reaped the completed
767 80. If a message was submitted locally by a user whose login name contained one
768 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
769 It caused a spool format error. I have fixed the spool reading code. A
770 related problem was that the "from" clause in the Received: line became
771 illegal because of the space(s). It is now covered by ${quote_local_part.
773 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
776 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
778 83. Patch from Sieve maintainer to fix unterminated string problem in
781 84. Some minor changes to the Linux configuration files to help with other
782 OS variants using glibc.
784 85. One more patch for Sieve to update vacation handling to latest spec.
787 ----------------------------------------------------
788 See the note above about the 4.44 and 4.50 releases.
789 ----------------------------------------------------
795 1. Change 4.43/35 introduced a bug that caused file counts to be
796 incorrectly computed when quota_filecount was set in an appendfile
799 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
800 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
803 3. Give more explanation in the error message when the command for a transport
804 filter fails to execute.
806 4. There are several places where Exim runs a non-Exim command in a
807 subprocess. The SIGUSR1 signal should be disabled for these processes. This
808 was being done only for the command run by the queryprogram router. It is
809 now done for all such subprocesses. The other cases are: ${run, transport
810 filters, and the commands run by the lmtp and pipe transports.
812 5. Some older OS have a limit of 256 on the maximum number of file
813 descriptors. Exim was using setrlimit() to set 1000 as a large value
814 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
815 systems. I've change it so that if it can't get 1000, it tries for 256.
817 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
818 was an oversight, and furthermore, ever since the addition of extra
819 controls (e.g. 4.43/32), the checks on when to allow different forms of
820 "control" were broken. There should now be diagnostics for all cases when a
821 control that does not make sense is encountered.
823 7. $recipients is now available in the predata ACL (oversight).
825 8. Tidy the search cache before the fork to do a delivery from a message
826 received from the command line. Otherwise the child will trigger a lookup
827 failure and thereby defer the delivery if it tries to use (for example) a
828 cached ldap connection that the parent has called unbind on.
830 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
831 of $address_data from the recipient verification was clobbered by the
834 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
835 was its contents. (It was OK if the option was not defined at all.)
837 11. A "Completed" log line is now written for messages that are removed from
838 the spool by the -Mrm option.
840 12. $host_address is now set to the target address during the checking of
843 13. When checking ignore_target_hosts for an ipliteral router, no host name was
844 being passed; this would have caused $sender_host_name to have been used if
845 matching the list had actually called for a host name (not very likely,
846 since this list is usually IP addresses). A host name is now passed as
849 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
850 code that specifies a non-restarting handler (typically sigaction() in
851 modern systems) in an attempt to fix a rare and obscure crash bug.
853 15. Narrowed the window for a race in the daemon that could cause it to ignore
854 SIGCHLD signals. This is not a major problem, because they are used only to
855 wake it up if nothing else does.
857 16. A malformed maildirsize file could cause Exim to calculate negative values
858 for the mailbox size or file count. Odd effects could occur as a result.
859 The maildirsize information is now recalculated if the size or filecount
862 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
863 support for a long time. Removed HAVE_SYS_VFS_H.
865 18. Updated exipick to current release from John Jetmore.
867 19. Allow an empty sender to be matched against a lookup in an address list.
868 Previously the only cases considered were a regular expression, or an
871 20. Exim went into a mad DNS lookup loop when doing a callout where the
872 host was specified on the transport, if the DNS lookup yielded more than
875 21. The RFC2047 encoding function was originally intended for short strings
876 such as real names; it was not keeping to the 75-character limit for
877 encoded words that the RFC imposes. It now respects the limit, and
878 generates multiple encoded words if necessary. To be on the safe side, I
879 have increased the buffer size for the ${rfc2047: expansion operator from
882 22. Failure to deliver a bounce message always caused it to be frozen, even if
883 there was an errors_to setting on the router. The errors_to setting is now
886 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
887 canonical form (fully expanded) before being placed in
888 $sender_host_address.
890 24. Updated eximstats to version 1.33
892 25. Include certificate and key file names in error message when GnuTLS fails
893 to set them up, because the GnuTLS error message doesn't include the name
894 of the failing file when there is a problem reading it.
896 26. Expand error message when OpenSSL has problems setting up cert/key files.
899 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
900 (this can affect the format of dates).
902 28. exim_tidydb, when checking for the continued existence of a message for
903 which it has found a message-specific retry record, was not finding
904 messages that were in split spool directories. Consequently, it was
905 deleting retry records that should have stayed in existence.
907 29. eximstats updated to version 1.35
908 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
909 1.35 - bugfix such that pie charts by volume are generated correctly
911 30. The SPA authentication driver was not abandoning authentication and moving
912 on to the next authenticator when an expansion was forced to fail,
913 contradicting the general specification for all authenticators. Instead it
914 was generating a temporary error. It now behaves as specified.
916 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
917 (the order specifies the preference for clients). The order is now AES256,
918 AES128, 3DES, ARCFOUR128.
920 31. Small patch to Sieve code - explicitly set From: when generating an
923 32. Exim crashed if a remote delivery caused a very long error message to be
924 recorded - for instance if somebody sent an entire SpamAssassin report back
925 as a large number of 550 error lines. This bug was coincidentally fixed by
926 increasing the size of one of Exim's internal buffers (big_buffer) that
927 happened as part of the Exiscan merge. However, to be on the safe side, I
928 have made the code more robust (and fixed the comments that describe what
931 33. Some experimental protocols are using DNS PTR records for new purposes. The
932 keys for these records are domain names, not reversed IP addresses. The
933 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
934 leaves it alone. Component reversal etc. now happens only for IP addresses.
937 34. The host_aton() function is supposed to be passed a string that is known
938 to be a valid IP address. However, in the case of IPv6 addresses, it was
939 not checking this. This is a hostage to fortune. Exim now panics and dies
940 if the condition is not met. A case was found where this could be provoked
941 from a dnsdb PTR lookup with an IPv6 address that had more than 8
942 components; fortuitously, this particular loophole had already been fixed
943 by change 4.50/55 or 4.44/33 above.
945 If there are any other similar loopholes, the new check in host_aton()
946 itself should stop them being exploited. The report I received stated that
947 data on the command line could provoke the exploit when Exim was running as
948 exim, but did not say which command line option was involved. All I could
949 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
953 35. There was a buffer overflow vulnerability in the SPA authentication code
954 (which came originally from the Samba project). I have added a test to the
955 spa_base64_to_bits() function which I hope fixes it.
958 36. The daemon start-up calls getloadavg() while still root for those OS that
959 need the first call to be done as root, but it missed one case: when
960 deliver_queue_load_max is set with deliver_drop_privilege. This is
961 necessary for the benefit of the queue runner, because there is no re-exec
962 when deliver_drop_privilege is set.
964 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
965 was in use, was not putting the data itself into the right store pool;
966 consequently, it could be overwritten for a subsequent message in the same
967 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
970 38. Sometimes the final signoff response after QUIT could fail to get
971 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
972 < 0 before doing a fflush(). This bug looks as though it goes back to the
973 introduction of TLS in release 3.20, but "sometimes" must have been rare
974 because the tests only now provoked it.
980 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
981 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
982 Now it returns an integer. A number of calls were still expecting a T/F
983 return. Fortuitously, in all cases, the tests worked in OK situations,
984 which is the norm. However, things would have gone wrong on any write
985 failures on the smtp file descriptor. This function is used when sending
986 messages over SMTP and also when doing verify callouts.
988 2. When Exim is called to do synchronous delivery of a locally submitted
989 message (the -odf or -odi options), it no longer closes stderr before doing
992 3. Implemented the mua_wrapper option.
994 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
996 5. Implemented the functions header_remove(), header_testname(),
997 header_add_at_position(), and receive_remove_recipient(), and exported them
1000 6. If an ACL "warn" statement specified the addition of headers, Exim already
1001 inserted X-ACL-Warn: at the start if there was no header name. However, it
1002 was not making this test for the second and subsequent header lines if
1003 there were newlines in the string. This meant that an invalid header could
1004 be inserted if Exim was badly configured.
1006 7. Allow an ACL "warn" statement to add header lines at the start or after all
1007 the Received: headers, as well as at the end.
1009 8. Added the rcpt_4xx retry error code.
1011 9. Added postmaster_mailfrom=xxx to callout verification option.
1013 10. Added mailfrom=xxxx to the callout verification option, for verify=
1016 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1017 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1019 12. Inserted some casts to stop certain compilers warning when using pointer
1020 differences as field lengths or precisions in printf-type calls (mostly
1021 affecting debugging statements).
1023 13. Added optional readline() support for -be (dynamically loaded).
1025 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1026 same clock tick as a message's arrival, so that its received time was the
1027 same as the "first fail" time on the retry record, and that message
1028 remained on the queue past the ultimate address timeout, every queue runner
1029 would try a delivery (because it was past the ultimate address timeout) but
1030 after another failure, the ultimate address timeout, which should have then
1031 bounced the address, did not kick in. This was a "< instead of <=" error;
1032 in most cases the first failure would have been in the next clock tick
1033 after the received time, and all would be well.
1035 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1036 being recognized when the domain list was tested by the match_domain
1037 condition in an expansion string.
1039 16. Added the ${str2b64: operator.
1041 17. Exim was always calling setrlimit() to set a large limit for the number of
1042 processes, without checking whether the existing limit was already
1043 adequate. (It did check for the limit on file descriptors.) Furthermore,
1044 errors from getrlimit() and setrlimit() were being ignored. Now they are
1045 logged to the main and panic logs, but Exim does carry on, to try to do its
1046 job under whatever limits there are.
1048 18. Imported PCRE 5.0.
1050 19. Trivial typo in log message " temporarily refused connection" (the leading
1053 20. If the log selector return_path_on_delivery was set and an address was
1054 redirected to /dev/null, the delivery process crashed because it assumed
1055 that a return path would always be set for a "successful" delivery. In this
1056 case, the whole delivery is bypassed as an optimization, and therefore no
1059 21. Internal re-arrangement: the function for sending a challenge and reading
1060 a response while authentication was assuming a zero-terminated challenge
1061 string. It's now changed to take a pointer and a length, to allow for
1062 binary data in such strings.
1064 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1066 23. Exim was not respecting finduser_retries when seeking the login of the
1067 uid under which it was called; it was always trying 10 times. (The default
1068 setting of finduser_retries is zero.) Also, it was sleeping after the final
1069 failure, which is pointless.
1071 24. Implemented tls_on_connect_ports.
1073 25. Implemented acl_smtp_predata.
1075 26. If the domain in control=submission is set empty, Exim assumes that the
1076 authenticated id is a complete email address when it generates From: or
1077 Sender: header lines.
1079 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1080 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1081 chown and chgrp in /bin and hostname in /usr/bin.
1083 28. Exim was keeping the "process log" file open after each use, just as it
1084 does for the main log. This opens the possibility of it remaining open for
1085 long periods when the USR1 signal hits a daemon. Occasional processlog
1086 errors were reported, that could have been caused by this. Anyway, it seems
1087 much more sensible not to leave this file open at all, so that is what now
1090 29. The long-running daemon process does not normally write to the log once it
1091 has entered its main loop, and it closes the log before doing so. This is
1092 so that log files can straightforwardly be renamed and moved. However,
1093 there are a couple of unusual error situations where the daemon does write
1094 log entries, and I had neglected to close the log afterwards.
1096 30. The text of an SMTP error response that was received during a remote
1097 delivery was being truncated at 512 bytes. This is too short for some of
1098 the long messages that one sometimes sees. I've increased the limit to
1101 31. It is now possible to make retry rules that apply only when a message has a
1102 specific sender, in particular, an empty sender.
1104 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1105 it possible to be selective about when SMTP synchronization is enforced.
1107 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1109 32. Implemented hosts_connection_nolog.
1111 33. Added an ACL for QUIT.
1113 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1116 35. Added mailbox_size and mailbox_filecount to appendfile.
1118 36. Added control = no_multiline_responses to ACLs.
1120 37. There was a bug in the logic of the code that waits for the clock to tick
1121 in the case where the clock went backwards by a substantial amount such
1122 that the microsecond fraction of "now" was more than the microsecond
1123 fraction of "then" (but the whole seconds number was less).
1125 38. Added support for the libradius Radius client library this is found on
1126 FreeBSD (previously only the radiusclient library was supported).
1132 1. When certain lookups returned multiple values in the form name=value, the
1133 quoting of the values was not always being done properly. Specifically:
1134 (a) If the value started with a double quote, but contained no whitespace,
1136 (b) If the value contained whitespace other than a space character (i.e.
1137 tabs or newlines or carriage returns) it was not quoted.
1138 This fix has been applied to the mysql and pgsql lookups by writing a
1139 separate quoting function and calling it from the lookup code. The fix
1140 should probably also be applied to nisplus, ibase and oracle lookups, but
1141 since I cannot test any of those, I have not disturbed their existing code.
1143 2. A hit in the callout cache for a specific address caused a log line with no
1144 reason for rejecting RCPT. Now it says "Previous (cached) callout
1145 verification failure".
1147 3. There was an off-by-one bug in the queryprogram router. An over-long
1148 return line was truncated at 256 instead of 255 characters, thereby
1149 overflowing its buffer with the terminating zero. As well as fixing this, I
1150 have increased the buffer size to 1024 (and made a note to document this).
1152 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1153 when Exim is waiting for an SMTP response from a remote server, Exim
1154 restarts its select() call on the socket, thereby resetting its timeout.
1155 This is not a problem when such interrupts are rare. Somebody set up a cron
1156 job to run exiwhat every 2 minutes, which is less than the normal select()
1157 timeout (5 or 10 minutes). This meant that the select() timeout never
1158 kicked in because it was always reset. I have fixed this by comparing the
1159 time when an interrupt arrives with the time at the start of the first call
1160 to select(). If more time than the timeout has elapsed, the interrupt is
1161 treated as a timeout.
1163 5. Some internal re-factoring in preparation for the addition of Sieve
1164 extensions (by MH). In particular, the "personal" test is moved to a
1165 separate function, and given an option for scanning Cc: and Bcc: (which is
1166 not set for Exim filters).
1168 6. When Exim created an email address using the login of the caller as the
1169 local part (e.g. when creating a From: or Sender: header line), it was not
1170 quoting the local part when it contained special characters such as @.
1172 7. Installed new OpenBSD configuration files.
1174 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1175 try to make them clearer.
1177 9. Callout options, other than the timeout value, were being ignored when
1178 verifying sender addresses in header lines. For example, when using
1180 verify = header_sender/callout=no_cache
1182 the cache was (incorrectly) being used.
1184 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1185 only the Cygwin environment.
1187 11. When return_path_on_delivery was set as a log selector, if different remote
1188 addresses in the same message used different return paths and parallel
1189 remote delivery occurred, the wrong values would sometimes be logged.
1190 (Whenever a remote delivery process finished, the return path value from
1191 the most recently started remote delivery process was logged.)
1193 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1194 lines when AUTH and/or TLS are in use. This is the "received protocol"
1195 field. Exim used to use "asmtp" for authenticated SMTP, without any
1196 indication (in the protocol name) for TLS use. Now it follows the RFC and
1197 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1198 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1199 names appear in log lines as well as in Received: header lines.
1201 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1202 extensions, and comparison tests, and to fix some bugs.
1204 14. Changes to the "personal" filter test:
1206 (1) The test was buggy in that it was just doing the equivalent of
1207 "contains" tests on header lines. For example, if a user's address was
1208 anne@some.where, the "personal" test would incorrectly be true for
1210 To: susanne@some.where
1212 This test is now done by extracting each address from the header in turn,
1213 and checking the entire address. Other tests that are part of "personal"
1214 are now done using regular expressions (for example, to check local parts
1215 of addresses in From: header lines).
1217 (2) The list of non-personal local parts in From: addresses has been
1218 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1219 taken from the Sieve specification recommendations.
1221 (3) If the message contains any header line starting with "List-" it is
1222 treated as non-personal.
1224 (4) The test for "circular" in the Subject: header line has been removed
1225 because it now seems ill-conceived.
1227 15. Minor typos in src/EDITME comments corrected.
1229 16. Installed latest exipick from John Jetmore.
1231 17. If headers_add on a router specified a text string that was too long for
1232 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1233 of string_sprintf() is now avoided.
1235 18. $message_body_size was not set (it was always zero) when running the DATA
1236 ACL and the local_scan() function.
1238 19. For the "mail" command in an Exim filter, no default was being set for
1239 the once_repeat time, causing a random time value to be used if "once" was
1240 specified. (If the value happened to be <= 0, no repeat happened.) The
1241 default is now 0s, meaning "never repeat". The "vacation" command was OK
1242 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1243 (I found it when inspecting the code).
1245 20. There is now an overall timeout for performing a callout verification. It
1246 defaults to 4 times the callout timeout, which applies to individual SMTP
1247 commands during the callout. The overall timeout applies when there is more
1248 than one host that can be tried. The timeout is checked before trying the
1249 next host. This prevents very long delays if there are a large number of
1250 hosts and all are timing out (e.g. when the network connections are timing
1251 out). The value of the overall timeout can be changed by specifying an
1252 additional sub-option for "callout", called "maxwait". For example:
1254 verify = sender/callout=5s,maxwait=20s
1256 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1257 to the end before writing, but this should make it even safer).
1259 22. Exim was forgetting that it had advertised PIPELINING for the second and
1260 subsequent messages on an SMTP connection. It was also not resetting its
1261 memory on STARTTLS and an internal HELO.
1263 23. When Exim logs an SMTP synchronization error within a session, it now
1264 records whether PIPELINING has been advertised or not.
1266 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1267 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1268 rather than long int.
1270 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1272 26. Added the never_mail option to autoreply.
1278 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1279 crash if the getsockname() call failed; this can happen if a connection is
1280 closed very soon after it is established. The problem was simply in the
1281 order in which certain operations were done, causing Exim to try to write
1282 to the SMTP stream before it had set up the file descriptor. The bug has
1283 been fixed by making things happen in the correct order.
1289 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1290 before the connection was closed, thus losing the rejection response.
1292 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1293 some early Solaris releases, but causes trouble in current releases where
1294 socklen_t is defined.
1296 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1299 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1302 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1303 that releasing the top of it at the end releases what was used for sub-
1304 expansions (unless the block got too big). However, discard this block if
1305 the first thing is a variable or header, so that we can use its block when
1306 it is dynamic (useful for very large $message_headers, for example).
1308 6. Lookups now cache *every* query, not just the most recent. A new, separate
1309 store pool is used for this. It can be recovered when all lookup caches are
1310 flushed. Lookups now release memory at the end of their result strings.
1311 This has involved some general refactoring of the lookup sources.
1313 7. Some code has been added to the store_xxx() functions to reduce the amount
1314 of flapping under certain conditions.
1316 8. log_incoming_interface used to affect only the <= reception log lines. Now
1317 it causes the local interface and port to be added to several more SMTP log
1318 lines, for example "SMTP connection from", and rejection lines.
1320 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1322 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1324 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1325 could be overwritten at the end of the current message (or the start of a
1326 new message if it was set in a HELO ACL). The value is now preserved for
1327 the duration of the SMTP connection.
1329 12. If a transport had a headers_rewrite setting, and a matching header line
1330 contained an unqualified address, that address was qualified, even if it
1331 did not match any rewriting rules. The underlying bug was that the values
1332 of the flags that permit the existence of unqualified sender and recipient
1333 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1334 non-local messages, and by -bnq for local messages) were not being
1335 preserved with the message after it was received.
1337 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1338 "next input=" as part of the text comprising the host identity instead of
1339 the correct text. The code was using the same buffer for two different
1340 strings. However, depending on which order the printing function evaluated
1341 its arguments, the bug did not always show up. Under Linux, for example, my
1342 test suite worked just fine.
1344 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1345 doesn't work with some older versions of Perl. It has been changed to "my",
1346 which in any case is probably the better facility to use.
1348 15. A really picky compiler found some instances of statements for creating
1349 error messages that either had too many or two few arguments for the format
1352 16. The size of the buffer for calls to the DNS resolver has been increased
1353 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1354 for addresses that have a lot of PTR records. This alleviates a problem; it
1355 does not fully solve it.
1357 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1358 buffer now truncates the list and logs the incident, which is the same
1359 action as happens when Exim is looking up a host name and its aliases.
1360 Previously in this situation something unpredictable would happen;
1361 sometimes it was "internal error: store_reset failed".
1363 18. If a server dropped the connection unexpectedly when an Exim client was
1364 using GnuTLS and trying to read a response, the client delivery process
1365 crashed while trying to generate an error log message.
1367 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1368 string, for example:
1370 warn message = H1: something\nH2: something
1372 the text was added as a single header line from Exim's point of view
1373 though it ended up OK in the delivered message. However, searching for the
1374 second and subsequent header lines using $h_h2: did not work. This has been
1375 fixed. Similarly, if a system filter added multiple headers in this way,
1376 the routers could not see them.
1378 20. Expanded the error message when iplsearch is called with an invalid key to
1379 suggest using net-iplsearch in a host list.
1381 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1382 ACLs are no longer actually imposed (and a message to that effect is
1385 22. If a "gecos" field in a passwd entry contained escaped characters, in
1386 particular, if it contained a \" sequence, Exim got it wrong when building
1387 a From: or a Sender: header from that name. A second bug also caused
1388 incorrect handling when an unquoted " was present following a character
1389 that needed quoting.
1391 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1392 was not being matched caselessly.
1394 24. Arranged for all hyphens in the exim.8 source to be escaped with
1397 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1398 properly. Recipient callout cache records were still being keyed to include
1399 the sender, even when use_sender was set false. This led to far more
1400 callouts that were necessary. The sender is no longer included in the key
1401 when use_sender is false.
1403 26. Added "control = submission" modifier to ACLs.
1405 27. Added the ${base62d: operator to decode base 62 numbers.
1407 28. dnsdb lookups can now access SRV records.
1409 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1410 the configuration file.
1412 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1413 (-v) mode. This makes the output for a verbose queue run more intelligible.
1415 31. Added a use_postmaster feature to recipient callouts.
1417 32. Added the $body_zerocount variable, containing the number of binary zero
1418 bytes in the message body.
1420 33. The time of last modification of the "new" subdirectory is now used as the
1421 "mailbox time last read" when there is a quota error for a maildir
1424 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1426 35. Added +ignore_unknown as a special item in host lists.
1428 36. Code for decoding IPv6 addresses in host lists is now included, even if
1429 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1430 address was recognized as an IP address, but was then not correctly decoded
1431 into binary, causing unexpected and incorrect effects when compared with
1438 1. Very minor rewording of debugging text in manualroute to say "list of
1439 hosts" instead of "hostlist".
1441 2. If verify=header_syntax was set, and a header line with an unqualified
1442 address (no domain) and a large number of spaces between the end of the
1443 name and the colon was received, the reception process suffered a buffer
1444 overflow, and (when I tested it) crashed. This was caused by some obsolete
1445 code that should have been removed. The fix is to remove it!
1447 3. When running in the test harness, delay a bit after writing a bounce
1448 message to get a bit more predictability in the log output.
1450 4. Added a call to search_tidyup() just before forking a reception process. In
1451 theory, someone could use a lookup in the expansion of smtp_accept_max_
1452 per_host which, without the tidyup, could leave open a database connection.
1454 5. Added the variables $recipient_data and $sender_data which get set from a
1455 lookup success in an ACL "recipients" or "senders" condition, or a router
1456 "senders" option, similar to $domain_data and $local_part_data.
1458 6. Moved the writing of debug_print from before to after the "senders" test
1461 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1462 problems for message scanning, either using a data ACL, or using
1463 local_scan() because the Received: header was not generated till after they
1464 were called (in order to set the time as the time of reception completion).
1465 I have revised the way this works. The header is now generated after the
1466 body is received, but before the ACL or local_scan() are called. After they
1467 are run, the timestamp in the header is updated.
1473 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1474 before starting a queue runner without re-exec. This happened only when
1475 deliver_drop_privilege was set or when the Exim user was set to root. The
1476 effect of the bug was that timeouts during subsequent deliveries caused
1477 crashes instead of being properly handled. The handler is now left at its
1478 default (and expected) setting.
1480 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1481 message, again when deliver_drop_privilege is set or Exim is run as root.
1482 The bug described in (1) was not present in this case, but the tidying up
1483 of the other signals was missing. I have made the two cases consistent.
1485 3. The ignore_target_hosts setting on a manualroute router was being ignored
1486 for hosts that were looked up using the /MX notation.
1488 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1491 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1492 operated on the sender address. After changing the $sender_address to <>
1493 for the sender address verify, Exim was re-instated it as the original
1494 (before rewriting) address, but remembering that it had rewritten it, so it
1495 wasn't rewriting it again. This bug also had the effect of breaking the
1496 sender address verification caching when the sender address was rewritten.
1498 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1499 This has been changed so that if the ip literal address matches
1500 ignore_target_hosts, the router declines.
1502 7. Added expansion conditions match_domain, match_address, and match_local_
1503 part (NOT match_host).
1505 8. The placeholder for the Received: header didn't have a length field set.
1507 9. Added code to Exim itself and to exim_lock to test for a specific race
1508 condition that could lead to file corruption when using MBX delivery. The
1509 issue is with the lockfile that is created in /tmp. If this file is removed
1510 after a process has opened it but before that process has acquired a lock,
1511 there is the potential for a second process to recreate the file and also
1512 acquire a lock. This could lead to two Exim processes writing to the file
1513 at the same time. The added code performs the same test as UW imapd; it
1514 checks after acquiring the lock that its file descriptor still refers to
1515 the same named file.
1517 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1518 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1521 11. Added the smtp_active_hostname option. If used, this will typically be made
1522 to depend on the incoming interface address. Because $interface_address is
1523 not set up until the daemon has forked a reception process, error responses
1524 that can happen earlier (such as "too many connections") no longer contain
1527 12. If an expansion in a condition on a "warn" statement fails because a lookup
1528 defers, the "warn" statement is abandoned, and the next ACL statement is
1529 processed. Previously this caused the whole ACL to be aborted.
1531 13. Added the iplsearch lookup type.
1533 14. Added ident_timeout as a log selector.
1535 15. Added tls_certificate_verified as a log selector.
1537 16. Added a global option tls_require_ciphers (compare the smtp transport
1538 option of the same name). This controls incoming TLS connections.
1540 17. I finally figured out how to make tls_require_ciphers do a similar thing
1541 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1542 before starting the TLS session.
1544 18. Tabs are now shown as \t in -bP output.
1546 19. If the log selector return_path_on_delivery was set, Exim crashed when
1547 bouncing a message because it had too many Received: header lines.
1549 20. If two routers both had headers_remove settings, and the first one included
1550 a superfluous trailing colon, the final name in the first list and the
1551 first name in the second list were incorrectly joined into one item (with a
1552 colon in the middle).
1558 1. Added -C and -D options to the exinext utility, mainly to make it easier
1559 to include in the automated testing, but these could be helpful when
1560 multiple configurations are in use.
1562 2. The exinext utility was not formatting the output nicely when there was
1563 an alternate port involved in the retry record key, nor when there was a
1564 message id as well (for retries that were specific to a specific message
1565 and a specific host). It was also confused by IPv6 addresses, because of
1566 the additional colons they contain. I have fixed the IPv4 problem, and
1567 patched it up to do a reasonable job for IPv6.
1569 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1570 delivery, the log line now contains "pipelined" if PIPELINING was used.
1572 4. An SMTP transport process used to panic and die if the bind() call to set
1573 an explicit outgoing interface failed. This has been changed; it is now
1574 treated in the same way as a connect() failure.
1576 5. A reference to $sender_host_name in the part of a conditional expansion
1577 that was being skipped was still causing a DNS lookup. This no longer
1580 6. The def: expansion condition was not recognizing references to header lines
1581 that used bh_ and bheader_.
1583 7. Added the _cache feature to named lists.
1585 8. The code for checking quota_filecount in the appendfile transport was
1586 allowing one more file than it should have been.
1588 9. For compatibility with Sendmail, the command line option
1596 and sets the incoming protocol and host name (for trusted callers). The
1597 host name and its colon can be omitted when only the protocol is to be set.
1598 Note the Exim already has two private options, -pd and -ps, that refer to
1599 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1600 "s", but I don't think that's a major issue.
1602 10. A number of refactoring changes to the code, none of which should affect
1605 (a) The number of logging options was getting close to filling up the
1606 32-bit word that was used as a bit map. I have split them into two classes:
1607 those that are passed in the argument to log_write(), and those that are
1608 only ever tested independently outside of that function. These are now in
1609 separate 32-bit words, so there is plenty of room for expansion again.
1610 There is no change in the user interface or the logging behaviour.
1612 (b) When building, for example, log lines, the code previously used a
1613 macro that called string_cat() twice, in order to add two strings. This is
1614 not really sufficiently general. Furthermore, there was one instance where
1615 it was actually wrong because one of the argument was used twice, and in
1616 one call a function was used. (As it happened, calling the function twice
1617 did not affect the overall behaviour.) The macro has been replaced by a
1618 function that can join an arbitrary number of extra strings onto a growing
1621 (c) The code for expansion conditions now uses a table and a binary chop
1622 instead of a serial search (which was left over from when there were very
1623 few conditions). Also, it now recognizes conditions like "pam" even when
1624 the relevant support is not compiled in: a suitably worded error message is
1625 given if an attempt is made to use such a condition.
1627 11. Added ${time_interval:xxxxx}.
1629 12. A bug was causing one of the ddress fields not to be passed back correctly
1630 from remote delivery subprocesses. The field in question was not being
1631 subsequently used, so this caused to problems in practice.
1633 13. Added new log selectors queue_time and deliver_time.
1635 14. Might have fixed a bug in maildirsizefile handling that threw up
1636 "unexpected character" debug warnings, and recalculated the data
1637 unnecessarily. In any case, I expanded the warning message to give more
1640 15. Added the message "Restricted characters in address" to the statements in
1641 the default ACL that block characters like @ and % in local parts.
1643 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1644 Three changes have been made:
1646 (a) There was a serious bug; a negative response to MAIL caused the whole
1647 recipient domain to be cached as invalid, thereby blocking all messages
1648 to all local parts at the same domain, from all senders. This bug has
1649 been fixed. The domain is no longer cached after a negative response to
1650 MAIL if the sender used is not empty.
1652 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1655 (c) A new callout option, "use_sender" has been added for people who want
1656 the modified behaviour.
1662 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1665 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1666 indeed breaks things for older releases.
1668 3. Added additional logging to the case where there is a problem reading data
1669 from a filter that is running in a subprocess using a pipe, in order to
1670 try to track down a specific problem.
1672 4. Testing facility fudge: when running in the test harness and attempting
1673 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1674 getting "No route to host". Convert this to a timeout.
1676 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1679 6. Some OS don't have socklen_t but use size_t instead. This affects the
1680 fifth argument of getsockopt() amongst other things. This is now
1681 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1682 can be set for individual OS. I have set it for SunOS5, OSF1, and
1683 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1684 some earlier ones do not.
1686 7. Change 4.30/15 was not doing the test caselessly.
1688 8. The standard form for an IPv6 address literal was being rejected by address
1689 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1690 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1691 this, as well as the form without the "IPv6" on the front (but only when
1692 address literals are enabled, of course).
1694 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1696 10. Exim crashed if a message with an empty sender address specified by -f
1697 encountered a router with an errors_to setting. This could be provoked only
1698 by a command such as
1702 where an empty string was supplied; "<>" did not hit this bug.
1704 11. Installed PCRE release 4.5.
1706 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1707 remained set. It is now erased.
1709 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1710 times from message ids (which are base 36 rather than the normal 62).
1712 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1713 were being counted as actual protocol errors, and logged if the log
1714 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1715 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1716 and DATA following a set of rejected RCPTs do not count as protocol errors.
1717 In other words, Exim assumes they were pipelined, though this may not
1718 actually be the case. Of course, in all cases the client gets an
1719 appropriate error code.
1721 15. If a lookup fails in an ACL condition, a message about the failure may
1722 be available; it is used if testing the ACL cannot continue, because most
1723 such messages specify what the cause of the deferral is. However, some
1724 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1725 that caused an old message to be retained and used if a later statement
1726 caused a defer, replacing the real cause of the deferral.
1728 16. If an IP address had so many PTR records that the DNS lookup buffer
1729 was not large enough to hold them, Exim could crash while trying to process
1730 the truncated data. It now detects and logs this case.
1732 17. Further to 4.21/58, another change has been made: if (and only if) the
1733 first line of a message (the first header line) ends with CRLF, a bare LF
1734 in a subsequent header line has a space inserted after it, so as not to
1735 terminate the header.
1737 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1738 unnecessarily, used atoi() instead of strtol(), and didn't check the
1739 termination when getting file sizes from file names by regex.
1741 19. Completely re-implemented the support for maildirsize files, in the light
1742 of a number of problems with the previous contributed implementation
1743 (4.30/29). In particular:
1745 . If the quota is zero, the maildirsize file is maintained, but no quota is
1748 . If the maildir directory does not exist, it is created before any attempt
1749 to write a maildirsize file.
1751 . The quota value in the file is just a cache; if the quota is changed in
1752 the transport, the new value overrides.
1754 . A regular expression is available for excluding directories from the
1757 20. The autoreply transport checks the characters in options that define the
1758 message's headers; it allows continued headers, but it was checking with
1759 isspace() after an embedded newline instead of explicitly looking for a
1762 21. If all the "regular" hosts to which an address was routed had passed their
1763 expiry times, and had not reached their retry times, the address was
1764 bounced, even if fallback hosts were defined. Now Exim should go on to try
1767 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1768 equivalent code in the SMTP transport. Some hosts send humungous responses
1769 to HELO/EHLO, more than 1024 it seems.
1771 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1772 gives compiler warnings in some environments. I've now done it "properly",
1775 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1776 (because of problems with the built-in one) was declared to return uschar *
1777 instead of char *, causing compiler failure.
1779 25. Fixed a file descriptor leak when processing alias/forward files.
1781 26. Fixed a minor format string issue in dbfn.c.
1783 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1785 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1786 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1789 29. When a sender address is verified, it is cached, to save repeating the test
1790 when there is more than one recipient in a message. However, when the
1791 verification involves a callout, it is possible for different callout
1792 options to be set for different recipients. It is too complicated to keep
1793 track of this in the cache, so now Exim always runs a verification when a
1794 callout is required, relying on the callout cache for the optimization.
1795 The overhead is duplication of the address routing, but this should not be
1798 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1799 to be verified with callout=postmaster, and the main callout worked but the
1800 postmaster check failed, the verification correctly failed. However, if a
1801 subsequent RCPT command asked for sender verification *without* the
1802 postmaster check, incorrect caching caused this verification also to fail,
1805 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1806 it was not caching the DNS options (qualify_single, search_parents) that
1807 were used when the lookup failed. A subsequent lookup with different
1808 options therefore always gave the same answer, though there were cases
1809 where it should not have. (Example: a "domains = !$mx_any" option on a
1810 dnslookup router: the "domains" option is always processed without any
1811 widening, but the router might have qualify_single set.) Now Exim uses the
1812 cached value only when the same options are set.
1814 32. Added John Jetmore's "exipick" utility to the distribution.
1816 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1817 than a timeout (e.g. a certificate is required, and is not provided), an
1818 Exim server now closes the connection immediately. Previously it waited for
1819 the client to close - but if the client is SSL, it seems that they each
1820 wait for each other, leading to a delay before one of them times out.
1822 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1823 maintained 0.8.x compatibility because I don't think many are using it, and
1824 it is clearly obsolete.
1826 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1829 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1830 client certificate was expired. A simple patch fixes this, though I don't
1831 understand the full logic of why the verify callback is called multiple
1834 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1835 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1836 which causes problems with some clients (such as the Certicom SSL Plus
1837 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1838 disables the coutermeasure allowing Eudora to connect."
1840 38. Exim was not checking that a write() to a log file succeeded. This could
1841 lead to Bad Things if a log got too big, in particular if it hit a file
1842 size limit. Exim now panics and dies if it cannot write to a log file, just
1843 as it does if it cannot open a log file.
1845 39. Modified OS/Makefile-Linux so that it now contains
1847 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1849 The two -D definitions ensure that Exim is compiled with large file
1850 support, which makes it possible to handle log files that are bigger than
1853 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1854 instance) a domain was checked against a named list that involved a lookup,
1855 causing $domain_data to be set, then another domain was checked against the
1856 same list, then the first domain was re-checked, the value of $domain_data
1857 after the final check could be wrong. In particular, if the second check
1858 failed, it could be set empty. This bug probably also applied to
1861 41. The strip_trailing_dot option was not being applied to the address given
1862 with the -f command-line option.
1864 42. The code for reading a message's header from the spool was incrementing
1865 $received_count, but never initializing it. This meant that the value was
1866 incorrect (doubled) while delivering a message in the same process in which
1867 it was received. In the most common configuration of Exim, this never
1868 happens - a fresh exec is done - but it can happen when
1869 deliver_drop_privilege is set.
1871 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1872 it now includes up to 150 characters of the unexpected data in the log
1875 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1876 and building data strings. The size of both of these buffers was 10 000
1877 bytes - far larger than anybody would *ever* want, thought I. Needless to
1878 say, somebody hit the limit. I have increased the maximum line length to
1879 20 000 and the maximum data length of concatenated lines to 100 000. I have
1880 also fixed two bugs, because there was no checking on these buffers. Tsk,
1881 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1882 buffer is too small.
1884 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1885 lsearch lookups. Now it does.
1887 46. When parsing a route_list item in a manualroute router, a fixed-length
1888 buffer was used for the list of hosts. I made this 1024 bytes long,
1889 thinking that nobody would ever have a list of hosts that long. Wrong.
1890 Somebody had a whole pile of complicated expansion conditions, and the
1891 string was silently truncated, leading to an expansion error. It turns out
1892 that it is easier to change to an unlimited length (owing to other changes
1893 that have happened since this code was originally written) than to build
1894 structure for giving a limitation error. The length of the item that
1895 expands into the list of hosts is now unlimited.
1897 47. The lsearch lookup could not handle data where the length of text line was
1898 more than 4095 characters. Such lines were truncated, leading to shortened
1899 data being returned. It should now handle lines of any length.
1901 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1902 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1905 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1906 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1907 no longer happens in comment lines. A list of replacements is now placed
1908 at the head of all of the source files, except those whose only change is
1909 to replace PERL_COMMAND in the very first #! line.
1911 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1912 messages on the queue, with a bottom-up merge sort, using code contributed
1913 by Michael Haardt. This should make operations like -bp somewhat faster on
1914 large queues. It won't affect queue runners, except when queue_run_in_order
1917 51. Installed eximstats 1.31 in the distribution.
1919 52. Added support for SRV lookups to the dnslookup router.
1921 53. If an ACL referred to $message_body or $message_body_end, the value was not
1922 reset for any messages that followed in the same SMTP session.
1924 54. The store-handling optimization for building very long strings was not
1925 differentiating between the different store pools. I don't think this
1926 actually made any difference in practice, but I've tidied it.
1928 55. While running the routers to verify a sender address, $sender_address
1929 was still set to the sender address. This is wrong, because when routing to
1930 send a bounce to the sender, it would be empty. Therefore, I have changed
1931 it so that, while verifying a sender address, $sender_address is set to <>.
1932 (There is no change to what happens when verifying a recipient address.)
1934 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1935 target A or AAAA records (if not already returned) without resetting the
1936 qualify_single or search_parents options of the DNS resolver. These are
1937 inappropriate in this case because the targets of MX and SRV records must
1938 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1939 target that, when qualified, matched something in the local domain. These
1940 two options are now turned off when doing these lookups.
1942 57. It seems that at least some releases of Reiserfs (which does not have the
1943 concept of a fixed number of inodes) returns zero and not -1 for the
1944 number of available inodes. This interacted badly with check_spool_inodes,
1945 which assumed that -1 was the "no such thing" setting. What I have done is
1946 to check that the total number of inodes is greater than zero before doing
1947 the test of how many are available.
1949 58. When a "warn" ACL statement has a log_message modifier, the message is
1950 remembered, and not repeated. This is to avoid a lot of repetition when a
1951 message has many recipients that cause the same warning to be written.
1952 Howewer, Exim was preserving the list of already written lines for an
1953 entire SMTP session, which doesn't seem right. The memory is now reset if a
1954 new message is started.
1956 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1957 debugging output unless log_rewrite was also set.
1959 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1960 of (int)(handle) when we know that handle contains (void *)(-1).
1962 61. The Exim daemon panic-logs an error return when it closes the incoming
1963 connection. However "connection reset by peer" seems to be common, and
1964 isn't really an error worthy of noting specially, so that particular error
1967 62. When Exim is trying to find all the local interfaces, it used to panic and
1968 die if the ioctl to get the interface flags failed. However, it seems that
1969 on at least one OS (Solaris 9) it is possible to have an interface that is
1970 included in the list of interfaces, but for which you get a failure error
1971 for this call. This happens when the interface is not "plumbed" into a
1972 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1973 failure of the "get flags" call assumes that the interface is down.
1975 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1976 makes life easier for people who are doing arithmetic on fields extracted
1977 from dates, where you often get leading zeros that should not be
1978 interpreted as octal.
1980 64. Added qualify_domain to the redirect router, to override the global
1983 65. If a pathologically long header line contained very many addresses (the
1984 report of this problem mentioned 10 000) and each of them was rewritten,
1985 Exim could use up a very large amount of memory. (It kept on making new
1986 copies of the header line as it rewrote, and never released the old ones.)
1987 At the expense of a bit more processing, the header rewriting function has
1988 been changed so that it no longer eats memory in this way.
1990 66. The generation of the Received: header has been moved from the time that a
1991 message starts to be received, to the time that it finishes. The timestamp
1992 in the Received: header should now be very close to that of the <= log
1993 line. There are two side-effects of this change:
1995 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1996 logged header lines no longer include the local Received: line, because
1997 it has not yet been created. The same applies to a copy of the message
1998 that is returned to a non-SMTP sender when a message is rejected.
2000 (b) When a filter file is tested using -bf, no additional Received: header
2001 is added to the test message. After some thought, I decided that this
2004 This change does not affect the value of $received_for. It is still set
2005 after address rewriting, but before local_scan() is called.
2007 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2009 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2010 gave an unhelpful panic error message, and a defer error. I have managed to
2011 change this behaviour so that it now rejects any supplied certificate,
2012 which seems right, as the list of acceptable certificates is empty.
2014 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2015 gave an unhelpful defer error. I have not managed to make this reject any
2016 supplied certificates, but the error message it gives is "no certificate
2017 supplied", which is not helpful.
2019 70. exigrep's output now also includes lines that are not associated with any
2020 message, but which match the given pattern. Implemented by a patch from
2021 Martin Sluka, which also tidied up the Perl a bit.
2023 71. Recipient callout verification, like sender verification, was using <> in
2024 the MAIL FROM command. This isn't really the right thing, since the actual
2025 sender may affect whether the remote host accepts the recipient or not. I
2026 have changed it to use the actual sender in the callout; this means that
2027 the cache record is now keyed on a recipient/sender pair, not just the
2028 recipient address. There doesn't seem to be a real danger of callout loops,
2029 since a callout by the remote host to check the sender would use <>.
2030 [SEE ABOVE: changed after hitting problems.]
2032 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2033 temporary errors. However, in the case of such a code being given after
2034 the end of a data transmission (i.e. after ".") Exim was failing to write
2035 a retry record for the message. (Yes, there was some broken host that was
2036 actually sending 8xx at this point.)
2038 73. An unknown lookup type in a host list could cause Exim to panic-die when
2039 the list was checked. (An example that provoked this was putting <; in the
2040 middle of a list instead of at the start.) If this happened during a DATA
2041 ACL check, a -D file could be left lying around. This kind of configuration
2042 error no longer causes Exim to die; instead it causes a defer errror. The
2043 incident is still logged to the main and panic logs.
2045 74. Buglet left over from Exim 3 conversion. The message "too many messages
2046 in one connection" was written to the rejectlog but not the mainlog, except
2047 when address rewriting (yes!) was being logged.
2049 75. Added write_rejectlog option.
2051 76. When a system filter was run not as root (that is, when system_filter_user
2052 was set), the values of the $n variables were not being returned to the
2053 main process; thus, they were not subsequently available in the $sn
2056 77. Added +return_path_on_delivery log selector.
2058 78. A connection timeout was being treated differently from recipients deferred
2059 when testing hosts_max_try with a message that was older than the host's
2060 retry timeout. (The host should not be counted, thus allowing all hosts to
2061 be tried at least once before bouncing.) This may have been the cause of an
2062 occasionally reported bug whereby a message would remain on the queue
2063 longer than the retry timeout, but would be bounced if a delivery was
2064 forced. I say "may" because I never totally pinned down the problem;
2065 setting up timeout/retry tests is difficult. See also the next item.
2067 79. The ultimate address timeout was not being applied to errors that involved
2068 a combination of host plus message (for example, a timeout on a MAIL
2069 command). When an address resolved to a number of possible hosts, and they
2070 were not all tried for each delivery (e.g. because of hosts_max_try), a
2071 message could remain on the queue longer than the retry timeout.
2073 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2076 81. Fixed an obscure SMTP outgoing bug which required at least the following
2077 conditions: (a) there was another message waiting for the same server;
2078 (b) the server returned 5xx to all RCPT commands in the first message so
2079 that the message was not completed; (c) the server dropped the connection
2080 or gave a negative response to the RSET that Exim sends to abort the
2081 transaction. The observed case was a dropped connection after DATA that had
2082 been sent in pipelining mode. That is, the server had advertised PIPELINING
2083 but was not implementing it correctly. The effect of the bug was incorrect
2084 behaviour, such as trying another host, and this could lead to a crash.
2090 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2091 and daemon.c were passed as pointers to ints; they should have been
2092 pointers to socklen_t variables (which are typically unsigned ints).
2094 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2097 3. Fixed a really odd bug that affected only the testing scheme; patching a
2098 certain fixed string in the binary changed the value of another string that
2099 happened to be identical to the end of the original first string.
2101 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2102 name", it returns that address as the IP address. On some operating
2103 systems (e.g. Solaris), it also passes back the IP address string as the
2104 "host name". However, on others (e.g. Linux), it passes back an empty
2105 string. Exim wasn't checking for this, and was changing the host name to an
2106 empty string, assuming it had been canonicized.
2108 5. Although rare, it is permitted to have more than one PTR record for a given
2109 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2110 all the names associated with an address, because they do in Solaris.
2111 However, it seems that they do not in Linux for data that comes from the
2112 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2113 I found this out when I moved to a new Linux workstation and tried to run
2114 the Exim test suite.
2116 To get round this problem I have changed the code so that it now does its
2117 own call to the DNS to look up PTR records when searching for a host name.
2118 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2119 addresses that are only in /etc/hosts are still found.
2121 This behaviour is, however, controlled by an option called host_lookup_
2122 order, which defaults to "bydns:byaddr". If people want to use the other
2123 order, or indeed, just use one or the other means of lookup, they can
2124 specify it in this variable.
2126 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2127 some operating systems, this comes back from gethostbyaddr() as an empty
2128 string, and this is what Exim used to test for. However, it seems that in
2129 other systems, "." is yielded. Exim now tests for this case too.
2131 7. The values of check_spool_space and check_log_space are now held internally
2132 as a number of kilobytes instead of an absolute number of bytes. If a
2133 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2134 kilobyte. This means that much larger values can be stored.
2136 8. Exim monitor: an attempt to get the action menu when not actually pointing
2137 at a message produces an empty menu entitled "No message selected". This
2138 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2139 no entries in it ("Shell widget menu has zero width and/or height"). So I
2140 have added a single, blank menu entry in this case.
2142 9. Added ${quote_local_part.
2144 10. MIME decoding is now applied to the contents of Subject: header lines when
2147 11. Now that a reference to $sender_host_address automatically causes a reverse
2148 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2149 host lookup before query-style lookups in lists that might use this
2150 variable. This has therefore been abolished, and the "net-" prefix is no
2151 longer necessary for query-style lookups.
2153 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2154 have been a typo for LFLAGS, so it has been changed.
2156 13. The install script calls Exim with "-C /dev/null" in order to find the
2157 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2158 to be output. Howeve, since Exim outputs its version number before the
2159 error, it didn't break the script. It just looked ugly. I fixed this by
2160 always allowing "-C /dev/null" if the caller is root.
2162 14. Ignore overlarge ACL variable number when reading spool file - insurance
2163 against a later release with more variables having written the file.
2165 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2166 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2167 this, as well as the form without the "IPv6" on the front.
2169 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2170 OS/Makefile-Darwin file.
2172 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2173 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2175 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2176 variable. However, some OS use ints for inodes. I've added cast to long int
2177 to get rid of the compiler warning.
2179 19. I had forgotten to lock out "/../" in configuration file names when
2180 ALT_CONFIG_PREFIX was set.
2182 20. Routers used for verification do not need to specify transports. However,
2183 if such a router generated a host list, and callout was configured, Exim
2184 crashed, because it could not find a port number from the (non-existent)
2185 transport. It now assumes port 25 in this circumstance.
2187 21. Added the -t option to exigrep.
2189 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2190 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2191 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2192 them is set, LOOKUP_LSEARCH is forced.
2194 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2196 24. Added sender and host information to the "rejected by local_scan()" log
2197 line; previously there was no indication of these.
2199 25. Added .include_if_exists.
2201 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2202 Linux. It turns out that not all file systems support this. Apparently some
2203 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2204 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2205 not supported on the file descriptor, is now ignored when Exim is trying to
2206 sync a directory. This applies only to Linux.
2208 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2210 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2211 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2214 29. Added support for maildirsize files from supplied patch (modified a bit).
2216 30. The use of :fail: followed by an empty string could lead Exim to respond to
2217 sender verification failures with (e.g.):
2219 550 Verification failed for <xxx>
2220 550 Sender verify failed
2222 where the first response line was missing the '-' that indicates it is not
2223 the final line of the response.
2225 31. The loop for finding the name of the user that called Exim had a hardwired
2226 limit of 10; it now uses the value of finduser_retries, which is used for
2227 all other user lookups.
2229 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2232 33. Exim was neglecting to zero errno before one call of strtol() when
2233 expanding a string and expecting an integer value. On some systems this
2234 resulted in spurious "integer overflow" errors. Also, it was casting the
2235 result into an int without checking.
2237 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2238 did not work. The code looks as if it has *never* worked, though it appears
2239 to have been documented since at least releast 1.62. I have made it work.
2241 35. The "timeout_DNS" error in retry rules, also documented since at least
2242 1.62, also never worked. As it isn't clear exactly what this means, and
2243 clearly it isn't a major issue, I have abolished the feature by treating it
2244 as "timeout", and writing a warning to the main and panic logs.
2246 36. The display of retry rules for -brt wasn't always showing the error code
2249 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2250 timeout_connect_A, timeout_connect_MX.
2252 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2253 to the empty sender.
2255 39. The daemon was not analysing the content of -oX till after it had closed
2256 stderr and disconnected from the controlling terminal. This meant that any
2257 syntax errors were only noted on the panic log, and the return code from
2258 the command was 0. By re-arranging the code a little, I've made the
2259 decoding happen first, so such errors now appear on stderr, and the return
2260 code is 1. However, the actual setting up of the sockets still happens in
2261 the disconnected process, so errors there are still only recorded on the
2264 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2265 connections (as happens on some IP stacks) was logged at start up time as
2266 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2267 it from "IPv6 and IPv4", which means that two separate sockets are being
2270 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2271 says whether AF_INET or AF_INET6 was passed as an argument.
2273 42. Exiwhat output was messed up when time zones were included in log
2276 43. Exiwhat now gives more information about the daemon's listening ports,
2277 and whether -tls-on-connect was used.
2279 44. The "port" option of the smtp transport is now expanded.
2281 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2282 silently ignored. Now an error message is written to the main and panic
2285 46. There's a new ACL modifier called "logwrite" which writes to a log file
2286 as soon as it is encountered.
2288 47. Added $local_user_uid and $local_user_gid at routing time.
2290 48. Exim crashed when trying to verify a sender address that was being
2293 49. Exim was recognizing only a space character after ".include". It now also
2294 recognizes a tab character.
2296 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2297 extracting the relevant information from the specification. The man page no
2298 longer contains scrambled data for the -d option, and I've added a section
2299 at the front about calling Exim under different names.
2301 51. Added "extra_headers" argument to the "mail" command in filter files.
2303 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2306 53. Installed eximstats 1.29.
2308 54. Added transport_filter_timeout as a generic transport option.
2310 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2311 Cc: header lines. This was required by RFC 822, but it not required by RFC
2314 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2315 incoming messages that did not have them. Now it does so only if the
2316 message originates locally, that is, if there is no associated remote host
2317 address. When Resent- header lines are present, this applies to the Resent-
2318 lines rather than the non-Resent- lines.
2320 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2321 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2323 58. Messages for configuration errors now include the name of the main
2324 configuration file - useful now that there may be more than one file in a
2325 list (.included file names were always shown).
2327 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2328 for those rare installations that do not start the daemon as root or run it
2329 setuid root. I've cut out the call to initgroups() if the daemon is not
2332 60. The Exim user and group can now be bound into the binary as text strings
2333 that are looked up at the start of Exim's processing.
2335 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2337 62. Added $mailstore_basename variable.
2339 63. Installed patch to sieve.c from Michael Haardt.
2341 64. When Exim failed to open the panic log after failing to open the main log,
2342 the original message it was trying to log was written to stderr and debug
2343 output, but if they were not available (the usual case in production), it
2344 was lost. Now it is written to syslog before the two lines that record the
2345 failures to open the logs.
2347 65. Users' Exim filters run in subprocesses under the user's uid. It is
2348 possible for a "deliver" command or an alias in a "personal" command to
2349 provoke an address rewrite. If logging of address rewriting is configured,
2350 this fails because the process is not running as root or exim. There may be
2351 a better way of dealing with this, but for the moment (because 4.30 needs
2352 to be released), I have disabled address rewrite logging when running a
2353 filter in a non-root, non-exim process.
2359 1. The buildconfig auxiliary program wasn't quoting the value set for
2360 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2361 not defined. This bug was masked in 4.22 by the effect that was fixed in
2364 2. Some messages that were rejected after a message id was allocated were
2365 shown as "incomplete" by exigrep. It no longer does this for messages that
2366 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2368 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2369 have allow_domain_literals set, the ID did not get logged in the <= line.
2370 Domain literals are now always recognized in Message-ID: header lines.
2372 4. The first argument for a ${extract expansion item is the key name or field
2373 number. Leading and trailing spaces in this item were not being ignored,
2374 causing some misleading effects.
2376 5. When deliver_drop_privilege was set, single queue runner processes started
2377 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2378 same command in the process it spins off) were not dropping privilege.
2380 6. When the daemon running as "exim" started a queue runner, it always
2381 re-executed Exim in the spun-off process. This is a waste of effort when
2382 deliver_drop_privilege is set. The new process now just calls the
2383 queue-runner function directly.
2389 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2392 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2393 ignored. Though the use of -oP was forcing the writing of a pid file, it
2394 was always written to the default place.
2396 3. If the message "no IP address found for host xxxx" is generated during
2397 incoming verification, it is now followed by identification of the incoming
2398 connection (so you can more easily find what provoked it).
2400 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2402 5. Added some features to "harden" Exim a bit more against certain attacks:
2404 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2405 be put in Local/Makefile. This is like the never_users runtime option,
2406 but it cannot be overridden. The default setting is "root".
2408 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2409 prefix string with which any file named in a -C command line option
2412 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2413 is retained for -C and -D only if the caller of Exim is root. Without
2414 it, the exim user may also use -C and -D and retain privilege.
2416 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2417 command line option is disabled.
2419 6. Macro names set by the -D option must start with an upper case letter, just
2420 like macro names defined in the configuration file.
2422 7. Added "dereference=" facility to LDAP.
2424 8. Two instances of the typo "uknown" in the source files are fixed.
2426 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2427 the Configure-Makefile script screwed up while processing it.
2429 10. Incorporated PCRE 4.4.
2431 11. The SMTP synchronization check was not operating right at the start of an
2432 SMTP session. For example, it could not catch a HELO sent before the client
2433 waited for the greeting. There is now a check for outstanding input at the
2434 point when the greeting is written. Because of the duplex, asynchronous
2435 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2436 way, but not yet received, when the check is performed.
2438 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2439 on TCP/IP sockets, because this apparently causes some broken clients to
2442 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2443 unchanged) from the Cygwin maintainer.
2445 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2446 was supported instead of testing for mbx. Effectively a typo.
2448 15. The spa authenticator server code was not checking that the input it
2449 received was valid base64.
2451 16. The debug output line for the "set" modifier in ACLs was not showing the
2452 name of the variable that was being set.
2454 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2456 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2457 lookup on its own. Something else was needed to trigger the lookup. For
2458 example, a match in host_lookup or the need for a host name in a host list.
2459 Now, if $sender_host_name is referenced and the host name has not yet been
2460 looked up, a lookup is performed. If the lookup fails, the variable remains
2461 empty, and $host_lookup_failed is set to "1".
2463 19. Added "eqi" as a case-independent comparison operator.
2465 20. The saslauthd authentication condition could segfault if neither service
2466 nor realm was specified.
2468 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2469 error message that was logged was misleading, and incoming SMTP
2470 connections were dropped. The message is now more accurate, and temporary
2471 errors are given to SMTP connections.
2473 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2474 (except RSET) with a 421 error, until QUIT is received. However, it was
2475 failing to send a response to QUIT.
2477 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2478 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2479 ACL is now run after the helo_try_verify_hosts code.
2481 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2482 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2483 case-independent, so other case variants are also recognized). Apparently
2484 some systems use these upper case variants.
2486 25. If more than two messages were waiting for the same host, and a transport
2487 filter was specified for the transport, Exim sent two messages over the
2488 same TCP/IP connection, and then failed with "socket operation on non-
2489 socket" when it tried to send the third.
2491 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2493 27. The extern definition of crypt16() in expand.c was not being excluded when
2494 the OS had its own crypt16() function.
2496 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2497 as a preferred synonym for return_size_limit, both as an option and as an
2500 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2502 30. Changed the default configuration ACL to relax the local part checking rule
2503 for addresses that are not in any local domains. For these addresses,
2504 slashes and pipe symbols are allowed within local parts, but the sequence
2505 /../ is explicitly forbidden.
2507 31. SPA server authentication was not clearing the challenge buffer before
2510 32. log_message in a "warn" ACL statement was writing to the reject log as
2511 well as to the main log, which contradicts the documentation and doesn't
2512 seem right (because no rejection is happening). So I have stopped it.
2514 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2515 However, I am unable to do any testing of this.
2517 34. Fixed an infelicity in the appendfile transport. When checking directories
2518 for a mailbox, to see if any needed to be created, it was accidentally
2519 using path names with one or more superfluous leading slashes; tracing
2520 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2522 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2523 contents are added to the log line that is written for every discarded
2524 recipient. (Previously a log_message setting was ignored.)
2526 36. The ${quote: operator now quotes the string if it is empty.
2528 37. The install script runs exim in order to find its version number. If for
2529 some reason other than non-existence or emptiness, which it checks, it
2530 could not run './exim', it was installing it with an empty version number,
2531 i.e. as "exim-". This error state is now caught, and the installation is
2534 38. An argument was missing from the function that creates an error message
2535 when Exim fails to connect to the socket for saslauthd authentication.
2536 This could cause Exim to crash, or give a corrupted message.
2538 39. Added isip, isip4, and isip6 to ${if conditions.
2540 40. The ACL variables $acl_xx are now saved with the message, and can be
2541 accessed later in routers, transports, and filters.
2543 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2544 strings in the file are not string-expanded.
2546 42. If a MAIL command specified a SIZE value that was too large to fit into an
2547 int variable, the check against message_size_limit failed. Such values are
2548 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2549 day this will have to be increased, but I don't think I want to be around
2550 when emails are that large.
2557 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2558 iconv() is not standard in FreeBSD.
2560 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2561 IPv6 enabled. The observed symptom was a segmentation fault on return from
2562 the function os_common_find_running_interfaces() in src/os.c.
2564 3. In the check_special_case() function in daemon.c I had used "errno" as an
2565 argument name, which causes warnings on some systems. This was basically a
2566 typo, since it was named "eno" in the comments!
2568 4. The code that waits for the clock to tick (at a resolution of some fraction
2569 of a second) so as to ensure message-id uniqueness was always waiting for
2570 at least one whole tick, when it could have waited for less. [This is
2571 almost certainly not relevant at current processor speeds, where it is
2572 unlikely to ever wait at all. But we try to future-proof.]
2574 5. The function that sleeps for a time interval that includes fractions of a
2575 second contained a race. It did not block SIGALRM between setting the
2576 timer, and suspending (a couple of lines later). If the interval was short
2577 and the sigsuspend() was delayed until after it had expired, the suspension
2578 never ended. On busy systems this could lead to processes getting stuck for
2581 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2582 process, before it forks any delivery processes. The open lookup caching
2583 mechanism meant that the open file or database connection was passed into
2584 the delivery process. The problem was that delivery processes always tidy
2585 up cached lookup data. This could cause a problem for the next delivery
2586 process started by the queue runner, because the external queue runner
2587 process does not know about the closure. So the next delivery process
2588 still has data in the lookup cache. In the case of a file lookup, there was
2589 no problem because closing a file descriptor in a subprocess doesn't affect
2590 the parent. However, if the lookup was caching a connection to a database,
2591 the connection was closed, and the second delivery process was likely to
2592 see errors such as "PGSQL: query failed: server closed the connection
2593 unexpectedly". The problem has been fixed by closing all cached lookups
2594 in a queue runner before running a delivery process.
2596 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2597 seem to have the "const" qualifier which it has on other OS. I've
2600 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2601 *of the same type* (client or server) with the same public name that an
2602 error should be diagnosed.
2604 9. When Exim looked up a host name for an IP address, but failed to find the
2605 original IP address when looking up the host name (a safety check), it
2606 output the message "<ip address> does not match any IP for NULL", which was
2607 confusing, to say the least. The bug was that the host name should have
2608 appeared instead of "NULL".
2610 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2611 user that is built into the binary, and the -C or -D options is used, root
2612 privilege is dropped before the configuration file is read. In addition,
2613 logging is switched to stderr instead of the normal log files. If the
2614 configuration then re-defines the Exim user, the unprivileged environment
2615 is probably not what is expected, so Exim logs a panic warning message (but
2618 However, if deliver_drop_privilege is set, the unprivileged state may well
2619 be exactly what is intended, so the warning has been cut out in that case,
2620 and Exim is allowed to try to write to its normal log files.
2626 1. smtp_return_error_details was not giving details for temporary sender
2627 or receiver verification errors.
2629 2. Diagnose a configuration error if two authenticators have the same public
2632 3. Exim used not to create the message log file for a message until the first
2633 delivery attempt. This could be confusing when incoming messages were held
2634 for policy or load reasons. The message log file is now created at the time
2635 the message is received, and an initial "Received" line is written to it.
2637 4. The automatically generated man page for command line options had a minor
2638 bug that caused no ill effects; however, a more serious problem was that
2639 the procedure for building the man page automatically didn't always
2640 operate. Consequently, release 4.20 contains an out-of-date version. This
2641 shouldn't happen again.
2643 5. When building Exim with embedded Perl support, the script that builds the
2644 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2645 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2647 6. The freeze_tell option was not being used for messages that were frozen on
2648 arrival, either by an ACL or by local_scan().
2650 7. Added the smtp_incomplete_transaction log selector.
2652 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2653 was accepting AUTH without a new EHLO.
2655 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2656 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2659 10. Logging of TCP/IP connections (when configured) now happens in the main
2660 daemon process instead of the child process, so that the TCP/IP connection
2661 count is more accurate (but it can never be perfect).
2663 11. The use of "drop" in a nested ACL was not being handled correctly in the
2664 outer ACL. Now, if condition failure induced by the nested "drop" causes
2665 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2666 or "require"), the connection is dropped.
2668 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2669 that yield "discard" can now be used with an "accept" or a "discard" verb,
2670 but an error is generated for any others (because I can't see a useful way
2671 to define what should happen).
2673 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2674 are now processed in the same way as lines in the Exim configuration file.
2675 In particular, continuation lines are supported.
2677 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2679 15. Added -ti meaning -t -i.
2681 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2682 domains, and warn by logging if others are found.
2684 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2685 returned from GIFCONF when Exim is trying to find the list of interfaces on
2686 a host. The code in os.c has been modified to copy each ifreq to an aligned
2687 structure in all cases.
2689 Also, in some cases, the returned ifreq's were being copied to a 'struct
2690 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2691 means the last couple of bytes of an IPv6 address could be chopped if the
2692 ifreq contained only a normal sockaddr (14 bytes storage).
2694 18. Named domain lists were not supported in the hosts_treat_as_local option.
2695 An entry such as +xxxx was not recognized, and was treated as a literal
2698 19. Ensure that header lines added by a DATA ACL are included in the reject log
2699 if the ACL subsequently rejects the message.
2701 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2702 MD5 (which is deprecated).
2704 21. When testing a filter file using -bf, Exim was writing a message when it
2705 took the sender from a "From " line in the message, but it was not doing so
2706 when it took $return_path from a Return-Path: header line. It now does.
2708 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2709 with a valid header line field name (a series of printing characters
2710 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2712 23. Changed "disc" in the source to "disk" to conform to the documentation and
2713 the book and for uniformity.
2715 24. Ignore Sendmail's -Ooption=value command line item.
2717 25. When execve() failed while trying to run a command in a pipe transport,
2718 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2719 could be confused with a return value of 69 from the command itself. This
2720 has been changed to 127, the value the shell returns if it is asked to run
2721 a non-existent command. The wording for the related log line suggests a
2722 non-existent command as the problem.
2724 26. If received_header_text expands to an empty string, do not add a Received:
2725 header line to the message. (Well, it adds a token one on the spool, but
2726 marks it "old" so that it doesn't get used or transmitted.)
2728 27. Installed eximstats 1.28 (addition of -nt option).
2730 28. There was no check for failure on the call to getsockname() in the daemon
2731 code. This can fail if there is a shortage of resources on the system, with
2732 ENOMEM, for example. A temporary error is now given on failure.
2734 29. Contrary to the C standard, it seems that in some environments, the
2735 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2736 program. Exim now does this explicitly; it affects the formatting of
2737 timestamps using strftime().
2739 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2740 complaints. I've now initialized all the variables, to avoid this.
2742 32. Header lines added by a system filter were not being "seen" during
2743 transport-time rewrites.
2745 33. The info_callback() function passed to OpenSSL is set up with type void
2746 (*)(SSL *, int, int), as described somewhere. However, when calling the
2747 function (actually a macro) that sets it up, the type void(*)() is
2748 expected. I've put in a cast to prevent warnings from picky compilers.
2750 34. If a DNS black list lookup found a CNAME record, but there were no A
2751 records associated with the domain it pointed at, Exim crashed.
2753 35. If a DNS black list lookup returned more than one A record, Exim ignored
2754 all but the first. It now scans all returned addresses if a particular IP
2755 value is being sought. In this situation, the contents of the
2756 $dnslist_value variable are a list of all the addresses, separated by a
2759 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2760 to accept a host name and all its aliases if the forward lookup for any of
2761 them yielded the IP address of the incoming connection. Now it accepts only
2762 those names whose forward lookup yields the correct IP address. Any other
2763 names are discarded. This closes a loophole whereby a rogue DNS
2764 administrator could create reverse DNS records to break through a
2765 wildcarded host restriction in an ACL.
2767 37. If a user filter or a system filter that ran in a subprocess used any of
2768 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2769 the wrong values were passed to the pipe command ($thisaddress had the
2770 value of $0, $0 had the value of $1, etc). This bug was introduced by
2771 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2773 38. Improved the line breaking for long SMTP error messages from ACLs.
2774 Previously, if there was no break point between 40 and 75 characters, Exim
2775 left the rest of the message alone. Two changes have been made: (a) I've
2776 reduced the minimum length to 35 characters; (b) if it can't find a break
2777 point between 35 and 75 characters, it looks ahead and uses the first one
2778 that it finds. This may give the occasional overlong line, but at least the
2779 remaining text gets split now.
2781 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2782 file descriptors might be low, and that setting 1000 would always raise it.
2783 It turns out that in some environments, the limit is already over 1000 and
2784 that lowering it causes trouble. So now Exim takes care not to decrease it.
2786 40. When delivering a message, the value of $return_path is set to $sender_
2787 address at the start of routing (routers may change the value). By an
2788 oversight, this default was not being set up when an address was tested by
2789 -bt or -bv, which affected the outcome if any router or filter referred to
2792 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2793 log only when "message" or "log_message" are set. However, if one of the
2794 conditions was an address verification, or a call to a nested ACL, the
2795 messages generated by the underlying test were being passed through. This
2796 no longer happens. The underlying message is available in $acl_verify_
2797 message for both "message" and "log_message" expansions, so it can be
2798 passed through if needed.
2800 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2801 new expansion $bh_ to give the encoded byte string without charset
2802 translation. Translation happens only if iconv() is available; HAVE_ICONV
2803 indicates this at build time. HEADERS_CHARSET gives the charset to
2804 translate to; headers_charset can change it in the configuration, and
2805 "headers charset" can change it in an individual filter file.
2807 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2808 that creates RFC 2047 encoded "words" labels them as that charset instead
2809 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2810 expansion operator; (ii) when Exim creates a From: line for a local
2811 message; (iii) when a header line is rewritten to include a "phrase" part.
2813 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2814 buggy, causing it to skip the first lines of messages whose message ID
2815 ended in 'D'. This would not have bitten before Exim release 4.14, because
2816 message IDs were unlikely to end in 'D' before then. The effect was to have
2817 incorrect size information for certain domains.
2819 45. #include "config.h" was missing at the start of the crypt16.c module. This
2820 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2823 46. If there was a timeout during a "random" callout check, Exim treated it as
2824 a failure of the random address, and carried on sending RSET and the real
2825 address. If the delay was just some slowness somewhere, the response to the
2826 original RCPT would be taken as a response to RSET and so on, causing
2827 mayhem of various kinds.
2829 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2830 when I implemented it. It didn't allow for the fact that some option values
2831 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2832 the right test for positive values.
2834 48. Domain names in DNS records are case-independent. Exim always looks them up
2835 in lower case. Some resolvers return domain names in exactly the case they
2836 appear in the zone file, that is, they may contain uppercase letters. Not
2837 all resolvers do this - some return always lower case. Exim was treating a
2838 change of case by a resolver as a change of domain, similar to a widening
2839 of a domain abbreviation. This triggered its re-routing code and so it was
2840 trying to route what was effectively the same domain again. This normally
2841 caused routing to fail (because the router wouldn't handle the domain
2842 twice). Now Exim checks for this case specially, and just changes the
2843 casing of the domain that it ultimately uses when it transmits the message
2846 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2849 50. If a filter generated a file delivery with a non-absolute name (possible if
2850 no home directory exists for the router), the forbid_file option was not
2853 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2854 the existing equality matching.
2856 52. Exim was using ints instead of ino_t variables in some places where it was
2857 dealing with inode numbers.
2859 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2860 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2861 called TMPDIR, and if it finds it is different, it changes its value.
2863 54. The smtp_printf() function is now made available to local_scan() so
2864 additional output lines can be written before returning. There is also an
2865 smtp_fflush() function to enable the detection of a dropped connection.
2866 The variables smtp_input and smtp_batched_input are exported to
2869 55. Changed the default runtime configuration: the message "Unknown user"
2870 has been removed from the ACL, and instead placed on the localuser router,
2871 using the cannot_route_message feature. This means that any verification
2872 failures that generate their own messages won't get overridden. Similarly,
2873 the "Unrouteable address" message that was in the ACL for unverifiable
2874 relay addresses has also been removed.
2876 56. Added hosts_avoid_esmtp to the smtp transport.
2878 57. The exicyclog script was not checking for the esoteric option
2879 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2880 will work only if exicyclog is run under the appropriate euid.
2882 58. Following a discussion on the list, the rules by which Exim recognises line
2883 endings on incoming messages have been changed. The -dropcr and drop_cr
2884 options are now no-ops, retained only for backwards compatibility. The
2885 following line terminators are recognized: LF CRLF CR. However, special
2886 processing applies to CR:
2888 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2889 nor a local message in the state where . is a terminator.
2891 (ii) If a bare CR is encountered in a header line, an extra space is added
2892 after the line terminator so as not to end the header. The reasoning
2893 behind this is that bare CRs in header lines are most likely either
2894 to be mistakes, or people trying to play silly games.
2896 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2897 was being incorrectly given as 18 bytes larger than it should have been.
2898 This is a VOB (very old bug).
2900 60. This may never have affected anything current, but just in case it has:
2901 When the local host is found other than at the start of a list of hosts,
2902 the local host, those with the same MX, and any that follow, are discarded.
2903 When the list in question was part of a longer list of hosts, the following
2904 hosts (not currently being processed) were also being discarded. This no
2905 longer happens. I'm not sure if this situation could ever has previously
2908 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2911 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2912 header. This is something that is recommended in a new Internet Draft, and
2913 is something that is documented as being done by Sendmail. There are two
2914 possible values. For messages generated by the autoreply transport, Exim
2917 Auto-Submitted: auto-replied
2919 whereas for all other generated messages (e.g. bounces) it adds
2921 Auto-Submitted: auto-generated
2923 63. The "personal" condition in filters now includes a test for the
2924 Auto-Submitted: header. If it contains the string "auto-" the message it
2925 not considered personal.
2927 64. Added rcpt_include_affixes as a generic transport option.
2929 65. Added queue_only_override (default true).
2931 66. Added the syslog_duplication option.
2933 67. If what should have been the first header line of a message consisted of
2934 a space followed by a colon, Exim was mis-interpreting it as a header line.
2935 It isn't of course - it is syntactically invalid and should therefore be
2936 treated as the start of the message body. The misbehaviour could have
2937 caused a number of strange effects, including loss of data in subsequent
2938 header lines, and spool format errors.
2940 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2941 client host had authenticated. This control can now be exercised by an ACL
2942 for more flexibility.
2944 69. By default, callouts do not happen when testing with -bh. There is now a
2945 variant, -bhc, which does actually run the callout code, including
2946 consulting and updating the callout cache.
2948 70. Added support for saslauthd authentication, courtesy of Alexander
2951 71. If statvfs() failed on the spool or log directories while checking their
2952 size for availability, Exim confusingly gave the error "space shortage".
2953 Furthermore, in debugging mode it crashed with a floating point exception.
2954 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2955 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2956 serious problem, Exim now writes to the main and panic logs when this
2957 happens, with details of the failure. It then refuses to accept the
2958 incoming message, giving the message "spool directory problem" or "log
2959 directory problem" with a 421 code for SMTP messages.
2961 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2962 0, 1, and 2 exist, because some OS complain for execs without them (see
2963 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2964 descriptors. However, the code omitted to check that the open succeeded,
2965 causing mysterious errors if for some reason the permissions on /dev/null
2966 got screwed. Now Exim writes a message to the main and panic logs, and
2967 bombs out if it can't open /dev/null.
2969 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2970 interact so that it is all more flexible. It is supposed to remain
2971 backwards compatible. Also added extra_local_interfaces.
2973 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2974 to bomb out with an assertion failure - to the client this appears as a
2975 connection drop. This problem occurs in the part of the code that was taken
2976 from the Samba project. Fortunately, the assertion is in a very simple
2977 function, so I have fixed this by reproducing the function inline in the
2978 one place where it is called, and arranging for authentication to fail
2979 instead of killing the process with assert().
2981 75. The SPA client code was not working when the server requested OEM rather
2982 than Unicode encoding.
2984 76. Added code to make require_files with a specific uid setting more usable in
2985 the case where statting the file as root fails - usually a non-root-mounted
2986 NFS file system. When this happens and the failure is EACCES, Exim now
2987 forks a subprocess and does the per-uid checking as the relevant uid.
2989 77. Added process_log_path.
2991 78. If log_file_path was not explicitly set, a setting of check_log_space or
2992 check_log_inodes was ignored.
2994 79. If a space check for the spool or log partitions fails, the incident is now
2995 logged. Of course, in the latter case the data may get lost...
2997 80. Added the %p formatting code to string_format() so that it can be used to
2998 print addresses in debug_print(). Adjusted all the address printing in the
2999 debugging in store.c to use %p rather than %d.
3001 81. There was a concern that a line of code in smtp_in.c could overflow a
3002 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3003 initially expressed, the concern was not well-founded, because trailing
3004 spaces are removed early. However, if the trailing spaces were followed by
3005 a NULL, they did not get removed, so the overflow was possible. Two fixes
3008 (a) I re-wrote the offending code in a cleaner fashion.
3009 (b) If an incoming SMTP command contains a NULL character, it is rejected
3012 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3013 runs initgroups(), so that if the Exim user is in any additional groups,
3014 they will be used during message reception.
3020 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob